IN RE: SOCIAL MEDIA ADOLESCENT ADDICTION/PERSONAL INJURY PRODUCTS LIABILITY LITIGATION (2024)

ORDER LARGELY DENYING IN PART META'S MOTION TO DISMISS THE MULTISTATE ATTORNEYS GENERAL COMPLAINT BUT LIMITING THE SCOPE OF CLAIMS; INCLUDING THE FLORIDA ATTORNEY GENERAL COMPLAINT; AND PERSONAL INJURY PLAINTIFFS' CONSUMER PROTECTION AND MISREPRESENTATION CLAIMS

This multi-district litigation (“MDL”) consolidates hundreds of actions brought on behalf of children and adolescents, school districts and local government entities, and state attorneys general (the “States”) alleging that several social media companies—Meta's ¹ Facebook and Instagram, Google's YouTube, ByteDance's TikTok, and Snapchat—designed their platforms to foster compulsive use by minors, resulting in a variety of harms to children, local governments, and the public health. Before the Court is a comprehensive motion to dismiss those claims rooted in notions of unfair practices, consumer-protection, misrepresentation, and concealment. For the reasons set forth in this Order, based on a careful review of the pleadings and the briefing submitted by the parties as well as oral argument heard on April 19, 2024, the Court GRANTS IN PART and DENIES IN PART the motion to dismiss.²

Much of the States' consumer protection claims are cognizable. Meta's alleged yearslong public campaign of deception as to the risks of addiction and mental harms to minors from platform use fits readily within these states' deceptive acts and practices framework. Meta's design, development, and deployment of certain product features plausibly constitutes unfair or unconscionable practices under all at-issue federal and state standards. However, as before, Section 230 provides a fairly significant limitation on these claims. Section 230 insulates the design and deployment of most features alleged to be unfair or unconscionable. Similarly, Section 230 protects against personal injury plaintiffs' consumer-protection, concealment, and misrepresentation theories to the same extent. That said, the Court declines to dismiss at this stage theories of liability predicated on a failure-to-warn of known risks of addiction attendant to any platform features or as to platform construction in general.

Given the nature of the action, this order allows claims to proceed but limits the scope, and therefore the motions to dismiss are generally denied.

A detailed summary of findings concludes this order.

I. BACKGROUND

A. Procedural Background

Thirty-four states, namely Arizona, California, Colorado, Connecticut, Delaware, Georgia, Hawai'i, Idaho, Illinois, Indiana, Kansas, Kentucky, Louisiana, Maine, Maryland, Michigan, Minnesota, Missouri, Nebraska, New Jersey, New York, North Carolina, North Dakota, Ohio, Oregon, Pennsylvania, Rhode Island, South Carolina, South Dakota, Virginia, Washington, West Virginia, and Wisconsin, through a consolidated complaint, and the state of Florida, through a separately filed complaint (collectively, the “States”),³ have sued several Meta entities, focusing on Meta's conduct in relation to its Facebook and Instagram social media platforms and their impact on young users.

As this MDL involves hundreds of actions brought by personal injury plaintiffs, school districts and local government entities, and state attorneys general, the Court has phased motion practice into multiple tracks based on the claim and claimant. (See Dkt. No. 451, Case Management Order No. 6 at 2–3.) Here, Meta moves to dismiss the following claims:

• All claims asserted in the complaint brought by the multistate attorney-general coalition. See Complaint (“Multistate Complaint”), Arizona, et al. v. Meta Platforms, Inc., et al., No. 4:23-cv-05448 (N.D. Cal. Oct. 24, 2023), Dkt. No. 1.

• All claims asserted in the complaint brought by the Florida attorney general. See Complaint (“Florida Complaint”), Office of the Attorney General, State of Florida, Department of Legal Affairs v. Meta Platforms, Inc., et al., No. 4:23-cv-05885 (N.D. Cal. Oct. 24, 2023), Dkt. No. 1.

• Counts 7, 8, and 9 asserted in the personal injury plaintiffs' second amended master complaint. See Second Amended Master Complaint (“2AMC”), In re: Social Media Adolescent Addiction/Personal Injury Prods. Liab. Litig., No. 4:22-md-3047 (N.D. Cal. Dec. 15, 2023), Dkt. No. 494.

The factual background underlying the personal injury plaintiffs' claims was discussed in the Court's order on the MDL defendants' first set of motions to dismiss. See In re Soc. Media Adolescent Addiction/Pers. Inj. Prod. Liab. Litig. (“Social Media I”), 702 F. Supp. 3d 809, 818–23 (N.D. Cal. 2023).⁴ This order is rooted in the relevant facts alleged by the States.

B. Relevant Facts Alleged

The States allege that Meta violated their consumer-protection laws by (1) designing and deploying platform features it knew were harmful to young users and (2) misleading and concealing from the public its knowledge of this harm. Third, the States also provide allegations specific to their claims under the Children's Online Privacy and Protection Act (“COPPA”), alleging that Meta impermissibly collects the data of children under 13 years old without obtaining prior consent from their parents. The Court briefly summarizes each of these three sets of allegations from the 217-page Multistate Complaint, which provides, in pertinent part:

1. Meta's Alleged Design and Deployment of Harmful Platform Features Targeting Young Users

Meta's core business model, as alleged, focuses on monetizing user information and attention on its platforms. (Multistate Compl. ¶ 54.) Meta then converts this user engagement into revenue obtained from advertisers, who can deploy targeted advertising based on the personal data Meta collects from each user. (Id. ¶ 55.) Thus, the more a user engages with Meta's platforms, the more revenue Meta can generate by collecting more user data and serving more targeted advertisements. (Id. ¶ 56.) On earnings calls, in federal filing reports, and in other public statements, Meta has correspondingly identified increasing user engagement and developing more effective targeted ad delivery as a key driver of revenue and top priority for future growth on its social media platforms. (Id. ¶¶ 57–63.) This motivation to increase and capture user engagement drives Meta's design of its platform features, in particular its recommendation algorithms. (E.g., id. ¶ 64.)

Meta's platforms contain a significant teen user-base. (See, e.g., id. ¶ 75 (“About 22 million teens log on to Instagram in the U.S. each day.”); id. ¶ 77 (“Within approximately two years of its purchase by Meta, over 50% of teenagers in the United States used Instagram ․”); id. ¶¶ 80–81 (providing Instagram's Adult Classifier Model's and Age Affinity Model's estimates of Instagrams millions of daily and monthly active users in age groups 13–17 and 18–23, broken down by state) Attracting and retaining young users thus allegedly forms a central focus of Meta's business development. (See, e.g., id. ¶ 68 (“As one Meta product designer summarized in an internal email, ‘[s]hort summary is the “the [sic] young ones are the best ones.” You want to bring people to your service young and early.’ ” (alterations in original)); id. ¶ 73 (According to an internal 2018 report, “Meta planned to ‘reverse the negative decline in teen engagement on Facebook’ by ‘focus[ing its] bets on early teens in markets where we have an acute teen problem (mostly the US & western markets) and will test new teen experiences for retention in the US first.’ ” (alteration in original)).)

In order to maintain and increase its young users' engagement with its platforms, Meta has allegedly developed and implemented features that it knows induce young users' extended, addictive, and compulsive use of social media platforms. (Id. ¶ 118.) Those designs, features, and functions include,⁵ but are not limited to:

Algorithmic Prioritization of Content. Meta's Instagram and Facebook platforms “employ Recommendation Algorithms that curate content from the main feeds and other parts of the Platforms.” (Id. ¶ 151.) Meta's algorithms rely on user signals (i.e., engagement with the platform) to “arrange each new piece of content to display to a user.” (Id. ¶ 152.) These algorithms were designed not to “benefit” the user but to maximize engagement by: (i) presenting material in an unpredictable sequence, relying on “variable reinforcement schedules,” a known psychological tactic (id. ¶¶ 155–56), (ii) harvesting data to “target user engagement on an individual level” (id. ¶ 166), and (iii) periodically presenting users with “emotionally gripping content ․ to provoke intense reactions” (e.g., relating to eating disorders, self-harm, suicide, violence, body-image issues, and more), a result of what Meta purportedly refers to as the algorithms' “preference amplification.” (Id. ¶ 176.) Despite Meta's representations to the contrary, this design results in harm to young users. (Id. ¶¶ 191–225.) Additionally, the dangers created by Meta's recommendation algorithms do not depend on the type of content suggested. Rather, the dopamine-manipulating “variable reinforcement schedules” (id. ¶¶ 156–61) on their own, irrespective of content, could still be harmful to young users by causing them to stay up late, leading to sleep deprivation, or by causing them to be distracted during school.

Infinite Scroll and Autoplay. Infinite scroll refers to the “partial display of additional content at the bottom of the user's screen, such that the user is typically unable to look at a single post in isolation.” (Id. ¶ 92.) This “teasing” of additional content continues indefinitely, as each time a user scrolls down their feed, new content is automatically loaded. (Id. ¶ 93.) Infinite scroll “makes it difficult for young users to disengage [from the content] because there is no natural end point for the display of new information.” (Id. ¶ 377.) Similarly, the “autoplay” feature refers to the continuous and automatic play of video content on Meta's platforms and “encourages young users to continuously engage on the Platform because it provides them with an ongoing supply of content.” (Id. ¶¶ 379, 381.)

Ephemeral Content. Meta made certain content available to users only temporarily, providing notifications and visual design cues which indicated “the content would soon disappear forever.” (Id. ¶ 97.) For example, in 2016 Meta introduced to Instagram a feature called “Stories,” through which users would upload images and narratives that remained on the platform for a short amount of time before disappearing. (Id. ¶ 99.) Meta designed such ephemeral content features to induce a sense of “FOMO” in young users, that is, a “Fear of Missing Out,” which would drive teen engagement. (Id. ¶¶ 98, 100.)

Notifications. Meta has introduced a variety of “push notifications,” which are “auditory and visual cues to alert users when accounts they follow add new content.” (Id. ¶ 95.) These notifications “allowed Instagram to draw its users back to the Platform at any time of day.” (Id. ¶ 96.) For instance, when a user goes “Live,” which is a feature giving “users the ability to livestream videos to followers or the public,” “the Instagram Platform sends out a notification.” (Id. ¶¶ 102, 105.) An internal memo from February 2017 noted that “of the 9.2 million broadcasts per day, ‘[Meta] found that 35% of [Live] broadcasters are teens (early and late high school).’ ” (Id. ¶ 106 (alterations in original).)⁶

Social Comparison Features. Meta's platforms contain features that “exacerbate social comparison, such as the quantification and display of “Like” counts on each piece of content on Instagram and Facebook.” (Id. ¶ 226.) “Likes,” for instance, provide a “way for users to express validation or approval of other users' photos or videos, by clicking or tapping a heart icon or the iconic thumbs-up icon.” (Id. ¶ 227.) According to the States, Meta is aware of the harms its social comparison features cause its young users. (See id. ¶ 229 (describing internal emails discussing the value of social comparison to Instagram's business model while causing harm to teen girls); id. ¶ 231 (“Specifically, Meta's researchers stated that they were ‘confident of a causal link between [seeing] Like counts and social comparison.’ ” (alteration in original)).)

“Multiple Accounts” Function. In 2016, Meta then deployed a feature on Instagram referred to as the “multiple accounts” function, which permits “users to register up to five accounts without having to log out of any one account to access another.” (Id. ¶ 370.) Consequently, “[t]een users with multiple accounts also have a higher probability of being exposed to harmful content.” (Id. ¶ 371.) Internal discussions from Meta developers discussed how the “multiple accounts” function, in tandem with its recommendation algorithms, would drive teen engagement on Instagram. (Id. ¶ 372.)

Filters. Meta has also developed features that promote and encourage eating disorders, focusing on “visual filters that simulate facial plastic surgery available to young users.” (Id. ¶¶ 333, 339.) Meta's leadership engaged in an internal debate regarding whether to extend the temporary ban instituted on this feature, given its propensity to promote “body image issues and anxiety among users and particularly among women and teenage girls,” as described in academic research. (Id. ¶ 349 (emphasis in original); id. ¶¶ 339–68.) One individual who agreed with the recommendation to extend the ban noted that “outside academics and experts consulted were nearly unanimous on the harm here.” (Id. ¶¶ 350–51.) Yet, Mark Zuckerberg, Meta's founder and Chief Executive Officer, vetoed the proposed permanent ban. (Id. ¶¶ 356–61.)

Features Relating to Restricting Time Spent on the Platforms. The States focus on two features relating to self-restricting time spent on Meta's platforms: “Daily Limit” and “Take a Break.” The “Daily Limit” feature “serves a pop-up notification whenever a user reaches the maximum amount of time they wish to spend on Instagram each day.” (Id. ¶ 579.) However, the feature was designed so the user can easily dismiss the notification and, contrary to its stated purpose, “regularly tempt[s]” users to revert to time-maximizing settings whenever the user reaches their chosen limit. (Id. ¶ 580.) The “Take a Break” tool “sends users a pop-up notification when they have spent more than a specified period of time scrolling without interruption.” (Id. ¶ 581.) This tool is also readily dismissible. (Id. ¶ 582.)

2. Meta's Alleged Awareness and Concealment of Harmful Platform Features From the Public

In parallel to Meta's development of these various platform features, Meta affirmatively misrepresented to the public the safety of its platforms while possessing knowledge of harms to young users. In addition to the examples outlined above specific to platform features, relevant here is Meta's general knowledge of harmful content and features, and affirmative statements regarding (i) the safety of Meta's platforms, (ii) Meta's under-13 account policy, (iii) the prevalence of harmful content on Meta's platforms, and (iv) Project Daisy.

Knowledge of Harmful Content and Features. Meta's internal research and reporting documents its knowledge of harms resulting from specific content on its platforms as well as harms related to the design of platform features. (See, e.g., Multistate Compl. ¶ 182 (“[A]n internal document from 2018 recognized that ‘seeing SSI [suicide and self-injury] admissions on Instagram is significantly associated with increased time spent’ on the Platform.” (alteration in original)); id. ¶ 195 (“[O]ne internal research paper stated” that “seeing more unconnected content [i.e., content from accounts that a user has not chosen to follow] is associated with worse appearance comparison․ In a recent listening session, one creator described Explore as ‘a landmine for everything I want to avoid on IG’ ․ because it triggers appearance comparison.” (second alteration in original)).)

Outside sources notified Meta of the harms to young users. For example, in “2019 and 2020, Mark Zuckerberg and Adam Mosseri met multiple times with Jonathan Haidt, a New York University professor studying the effects of social media on teens' mental health. Haidt emphasized to Zuckerberg his concerns regarding Meta's Platforms and its effects on ‘teen girls, whose rates of depression and self-harm have increased the most.’ ” (Id. ¶ 418.) Haidt apparently conducted independent research as to the effects of Meta's platforms, detailing how, for example, “in 2013 alone—the year after Instagram's surge in popularity among young users—the suicide rate for 13-year-old girls jumped by around 50%.” (Id. ¶¶ 508 & n.14, 525 & n.32.)

Public Statements of Safety. The Multistate Complaint collects numerous public statements made by Meta and its leadership in a variety of forums likewise varying in character and specificity as relates to the safety of its platforms. A few examples include:

• “In response to a ‘60 Minutes’ exposé on Meta's Platforms and the harms they cause in October 2021, Meta itself prepared the following public statement: ‘Protecting our community is more important than maximizing our profits.’ ” (Id. ¶ 127.)

• “Instagram's website characterized the Instagram app as a ‘safe and supportive community’ for its users.” (Id. ¶ 130.)

• In January 2018, Zuckerberg said the company was “focused on making sure Facebook isn't just fun to use, but also good for people's wellbeing,” as reported by the Guardian. (Id. ¶ 222.)

• “In June 2019, Mosseri (Head of Instagram) told CBS in an interview that teen well-being is a top priority. And two years later, in May 2021, Mosseri minimized Instagram's negative impact on teens, characterizing it to reporters as ‘quite small,’ as reported by the Wall Street Journal that September.” (Id. ¶ 126.)

• “On September 30, 2021, Meta executive Antigone Davis testified to Congress, ‘[w]e have put in place multiple protections to create safe and age-appropriate experiences for people between the ages of 13 and 17.’ ” (Id. ¶ 128.)

• “Meta has also claimed, in a statement published by Gizmodo on October 3, 2021, to ‘do internal research to ask hard questions and find out how we can best improve the experience for teens.’ ” (Id. ¶ 137.)

• “The Instagram website also boasts that ‘[a]t Instagram, we have guidelines that govern what content we recommend to people’ and specifies that Instagram ‘avoid[s] making recommendations that may be inappropriate for younger viewers ․ We use technology to detect both content and accounts that don't meet these Recommendations Guidelines and to help us avoid recommending them. As always, content that goes against our Community Guidelines will be removed from Instagram.’ ” (Id. ¶ 188.)

Under-13 Account Policy. “Despite its public-facing claims that users under the age of 13 are not allowed on Instagram” (id. ¶ 648), Meta effectively adheres to a contrary policy. “As an internal Meta document from 2018 acknowledges: ‘we do very little to keep U13s off our platform.’ ” (Id. ¶ 647.) “In practice, Meta employees often do not take action unless they can verify that the account actually belongs to an underage user. This results in underage accounts remaining on the Platform despite having been reported to Meta as belonging to users under the age of 13. As a matter of policy, Meta employees generally do not take action if the reported account does not contain a user bio or photos.” (Id. ¶ 686.) For instance, on one occasion, Meta allegedly chose not to remove four under-13 accounts when a parent provided explicit notice that the four accounts were created by her 12-year-old daughter and requested those accounts be removed. (Id. ¶ 668 (“[E]mployees concluded that ‘the accounts were ignored’ in part because representatives of Meta ‘couldn't tell for sure the user was underage.’ ”)

Prevalence of Harmful Content. In general, “the incidence or prevalence of negative or harmful user experiences on” Instagram and Facebook were deceptive. (Id. ¶ 846(c).) The States focus on Meta's Community Standard Enforcement Reports (“CSER”), which “describe the percentage of content posted on Instagram and Facebook that Meta removes for violating Instagram and Facebook's Community Standards or Guidelines” (id. ¶ 461), alleging the CSER omits relevant information from Meta's Tracking Reach of Integrity Problems Survey (“TRIPS”) and Bad Experiences & Encounters Framework (“BEEF”) reports. For example, while Meta's CSER from the third quarter of 2021 “concluded that only ‘0.05-0.06%’ of views on Instagram were of content that violated Meta's standards on bullying and harassment,” contemporaneous internal user survey data indicated that 28.3% of surveyed Instagram users “witnessed bullying on the platform within the last seven days.” (Id. ¶¶ 490–91.) The same CSER report estimated that “less than 0.05% of views were of content that violated [Meta's] standards against Suicide & Self-Injury,” whereas “contemporaneous internal survey data showed that during 2021, 6.7% of surveyed Instagram users had seen self-harm content within the last seven days” and of “users between 13 and 15 years of age, 8.4% had seen content relating to self-harm on Instagram within the last seven days.” (Id. ¶¶ 481–82.)

Project Daisy. Finally, Project Daisy was an internal Meta research experiment conducted in 2020. “Meta carried out two pilot versions of Project Daisy: Pure Daisy (wherein the Like counts on all posts except one's own were hidden) and Popular Daisy (wherein the Like counts on posts from certain highly followed accounts were visible, but the Like counts on the average users' posts were hidden).” (Id. ¶ 241.) Both programs reduced the negative impact of seeing posts with many likes (id. ¶ 242), that is, “hiding Like counts [on Instagram posts] led to decreases in negative social comparison” (id. ¶ 240 (alteration in original)). Researchers “recommended hiding Like counts for the entire Instagram Platform.” (Id.) Outside experts allegedly told Meta they should maintain Daisy as a default for teens. (Id. ¶ 262.) Internal researchers agreed. (Id. ¶¶ 261–62 (Meta researchers “HUGELY pushed” for Project Daisy as a “default for teens.”); id. ¶ 290 (Project Daisy “is one of the clearest things (supported by research) that we can do to positively impact social comparison and well-being on” Instagram).)

Meta abandoned the project, with leadership conflicted over its demise, despite Adam Mosseri (head of Instagram) having publicly promised that Meta would implement Daisy. (Id. ¶¶ 262–65.) Ultimately, Project Daisy was projected to damage revenue (id. ¶ 250) and got “stuck in a political war” among Meta's leadership (id. ¶ 264). In public statements, Meta represented that Project Daisy was not implemented because it was “beneficial for some, and annoying to others,” and that it “ultimately didn't have as strong an impact as we'd hoped,” among other statements. (Id. ¶¶ 284–87.)

3. COPPA Allegations

The Multistate Complaint also provides that (i) Meta has actual knowledge of under-13 users on its platforms, (ii) Meta's platforms are “directed to children,” under the COPPA Rule, and (iii) Meta collects the personal information of under-13 users without appropriate notice or parental consent. (See Multistate Compl. ¶¶ 631–835.)

The Court need not exhaustively describe the States' COPPA allegations at this stage, but instead makes three observations. First, as to actual knowledge of under-13 users, internal records suggest both Instagram and Facebook assess the demographic make-up of the platforms' users, with explicit acknowledgement of an under-13 user base. (See, e.g., id. ¶ 646 (referring to “charts boasting Instagram's penetration into 11-and 12-year-old demographic cohorts”); id. ¶ 652 (“Th[is] chart reveals that in 2016, approximately 20% of users from the 2004 birth year cohort (i.e., Instagram users who were 12 years old at that time) used Instagram on at least a daily basis.”); id. ¶¶ 657–58 (noting Zuckerberg received a report which “explained that Meta ‘can estimate that there were 4M[illion] people under 13 in 2015 on IG [in the US]. This represents around 30% of all 10-12 year[ ] old[s] in the US.’ ” (alterations in original)); id. ¶ 818 (“Meta possesses data from 2020 indicating that, out of 3,989 children surveyed, 31% of child respondents aged 6-9 and 44% of child respondents aged 10 to 12-years-old had used Facebook.”); id. ¶ 821 (“Upon information and belief, Meta has confirmed its knowledge of specific under-13 user accounts through its review of data generated by Meta's age-estimation algorithms confirming that millions of individual Facebook accounts belong to children under the age of 13.”).)

Second, to argue that Meta's platforms are “directed to children,” the States rely on Meta's own advertising campaigns of the platforms, which for example “featur[e] actors who appear to be children or teens,” as well as third-party content hosted on the platforms. (See, e.g., id. ¶¶ 759–66 (discussing advertisements on Instagram); id. ¶¶ 794–98 (discussing purportedly child-directed, third-party content on Instagram); id. ¶¶ 828–32 (discussing advertisements and third-party content on Facebook); id. ¶ 828 (“[T]housands of Facebook pages and accounts are child-oriented, including because they feature child-oriented subject matter, characters, activities, and music, as well as child models, child celebrities, and celebrities who appeal to children.”).)

Third, the States allege three violations of COPPA, each of which “constitutes an unfair or deceptive act or practice, in violation of 15 U.S.C. § 45” of the Federal Trade Commission (“FTC”) Act (Multistate Compl. ¶ 856),⁷ namely:

• Failure to Notify Parents “․ about the information [Meta] collects from children and how it uses such information, and its disclosure practices are in violation of Sections 312.4(b) and 312.4(c) of the COPPA Rule, 16 C.F.R. § 312.4(b)-312.4(c).” (Id. ¶ 853; see also Florida Am. Compl. ¶ 145.)

• Failure to Provide Platform Notice “․ on its Social Media Platforms about the information it collects from children and how it uses such information, and its disclosure practices are in violation of Section 312.4(d) of the COPPA Rule, 16 C.F.R. § 312.4(d).” (Multistate Compl. ¶ 854; see also Florida Am. Compl. ¶ 146.)

• Failure to Obtain Parental Consent “․ prior to collecting or using any personal information of children, in violation of Section 312.5 of the COPPA Rule, 16 C.F.R. § 312.5.” (Multistate Compl. ¶ 855; see also Florida Am. Compl. ¶ 147.)

C. Theories of Liability

Upon these summarized facts, the States allege numerous violations of their consumer-protection laws across fifty-four counts. These claims fall into two buckets: claims of (i) deceptive acts and practices and (ii) unfair and/or unconscionable acts and practices.

1. Deceptive Acts and Practices

Paragraph 846 of the Multistate Complaint instructs on the nature of the alleged deceptive acts and practices. There, the States claim that “Meta misrepresented, directly or indirectly, expressly or by implication that”:

(a) “its Social Media Platforms are not psychologically or physically harmful for young users and are not designed to induce young users' compulsive and extended use, when they are in fact so designed” (Multistate Compl. ¶ 846(a));

(b) “its Social Media Platforms are less addictive and/or less likely to result in psychological and physical harm for young users than its Social Media Platforms are in reality” (id. ¶ 846(b));

(c) “through the publication of CSER reports and intentional omission of material BEEF and TRIPS data from those reports, and through other communications, that the incidence or prevalence of negative or harmful user experiences on Meta's Social Media Platforms was lower than it actually was” (id. ¶ 846(c));

(d) “it prioritized young users' health and safety over maximizing profits, when in fact Meta subordinated young user health and safety to its goal of maximizing profits by prolonging young users' time spent on its Social Media Platforms” (id. ¶ 846(d));

(e) “Meta prevents under-13 users from using Instagram and/or Facebook when in fact Meta was aware that it does not prevent under-13 users from using Instagram and Facebook” (id. ¶ 846(e)); and

(f) “Meta's collection of user data was not for the purpose of causing those users to become addicted to the Social Media Platforms, when in reality that was one of the purposes for which Meta collected user data” (id. ¶ 846(f)).8

Two genres emerge: either Meta committed the misrepresentation by affirmative statement, or “indirectly ․ by implication,” i.e., a misrepresentation by omission.

2. Unfair and/or Unconscionable Acts and Practices

Paragraph 847 of the Multistate Complaint, reproduced below, summarizes the States' allegations of unfair and/or unconscionable acts and practices:

(a) “Meta targeted its Social Media Platforms to young users while knowingly designing its Social Media Platforms to include features that Meta knew to be psychologically and physically harmful to young users—including features known to promote compulsive, prolonged, and unhealthy use by young users.” (Multistate Compl. ¶ 847(a).)

(b) “Meta utilized Social Media Platform features that unfairly and/or unconscionably harm young users independently of any actions taken by third-party users of Meta's Platforms. These features include infinite scroll, ephemeral content features, autoplay, quantification and display of ‘Likes,’ and disruptive alerts, all of which were unfairly and/or unconscionably utilized by Meta to extract additional time and attention from young users whose developing brains were not equipped to resist those manipulative tactics.” (Id. ¶ 847(b).)

(c) “Meta designed, developed, and deployed disruptive audiovisual and vibration notifications and alerts and ephemeral content features in a way that unfairly and/or unconscionably exploited young users' psychological vulnerabilities and cultivated a sense of ‘fear of missing out’ in order to induce young users to spend more time than they would otherwise choose on Meta's Social Media Platforms.” (Id. ¶ 847(c).)

(d) “Meta algorithmically served content to young users, according to ‘variable reinforcement schedules,’ thereby manipulating dopamine releases in young users, unfairly or unconscionably inducing them to engage repeatedly with its products—much like a gambler at a slot machine.” (Id. ¶ 847(d).)

(e) “Meta collected the personal information of under-13 users of Instagram and Facebook without first obtaining verifiable parental consent, which violated COPPA and the COPPA Rule.” (Id. ¶ 847(e).)

Here, the States target the design, development, and deployment of features of Meta's platforms rather than any representations or omissions.

II. LEGAL FRAMEWORK

In an MDL, the transferee court applies the law of its circuit to issues of federal law, but on issues of state law it applies the state law that would have been applied to the underlying case as if it had never been transferred into the MDL. In re Anthem, Inc. Data Breach Litig., 2015 WL 5286992, at *2 (N.D. Cal. Sept. 9, 2015) (collecting Ninth Circuit cases).

The standard under Federal Rule of Civil Procedure 12(b)(6) is well-known and not in dispute. “To survive a motion to dismiss for failure to state a claim after the Supreme Court's decisions in Iqbal and Twombly, plaintiffs' allegations must suggest that their claim has at least a plausible chance of success.” Levitt v. Yelp! Inc., 765 F.3d 1123, 1134–35 (9th Cir. 2014) (cleaned up). The district court must assume that the plaintiffs' allegations are true and draw all reasonable inferences in their favor. The court need not, however, construe as true conclusory statements or unreasonable inferences. In re Gilead Scis. Sec. Litig., 536 F.3d 1049, 1055 (9th Cir. 2008). These well-established standards apply with equal force in MDL proceedings. See In re Optical Disk Drive Antitrust Litig., 2011 WL 3894376, at *8–9 (N.D. Cal. Aug. 3, 2011) (applying such standard in the context of an MDL); In re Zofran (Ondansetron) Prod. Liab. Litig., 2017 WL 1458193, at *5 (D. Mass. Apr. 24, 2017) (the “creation of an MDL proceeding does not suspend [or change] the requirements of the Federal Rules of Civil Procedure”).

The claims at issue raise multiple broad and distinct theories of harm regarding a wide variety of alleged conduct by Meta and, to a limited extent, other defendants (see Section VI.B). In the interest of efficiency and clarity, this Order is organized as follows:

• Section III COPPA Analysis

• Section IV Section 230

• Section V.A Consumer Protection Claims: Deceptive Acts and Practices

• Section V.B Consumer Protection Claims: Unconscionable and/or Unfair Acts and Practices

• Section V.C Consumer Protection Claims: Statutory Scope

• Section V.D Consumer Protection Claims: Restitution

• Section VI Personal Injury Plaintiffs' Similar Claims

III. THE CHILDREN'S ONLINE PRIVACY PROTECTION ACT

Meta brings a partial motion to dismiss the States' claims under COPPA, i.e., the Children's Online Privacy Protection Act of 1998, 15 U.S.C. §§ 6501–6505.⁹ By way of background, COPPA aims to “safeguard the confidentiality, security, and integrity of ․ children's personal online information” by requiring “companies that operate websites and online services marketed toward children ․ [to] provide certain disclosures about their data collection activities.” Jones v. Google LLC, 73 F.4th 636, 641 (9th Cir. 2023). Under COPPA it is unlawful for a website or online service that is “directed to children” or has “actual knowledge” of child users to collect personal information from those children unless the platform collects information in a manner permitted by FTC regulations. 15 U.S.C. § 6502(a)(1)). COPPA further defines a “website or online service directed to children” as “(i) a commercial website or online service that is targeted to children; or (ii) that portion of a commercial website or online service that is targeted to children.” 15 U.S.C. § 6501(10)(A). Thus, a company can become subject to COPPA's requirements in two ways: if its platform is “directed to children” or it has “actual knowledge” of child users.

Here, Meta moves solely on the grounds that its Facebook and Instagram platforms are not “directed to children” and thus they need not abide by COPPA's requirements under that theory. Meta has conceded, for the purposes of this motion only, that the States have adequately alleged it had actual knowledge under the statute that children under 13 were using its platforms during the relevant time period. Thus, under this theory the claim will proceed. Meta's motion to dismiss is DENIED as to the States' COPPA claim. Motions to dismiss should not be used for piecemeal purposes. See Dermansky v. Young Turks, Inc., 2023 WL 8884364, at *2 (C.D. Cal. Nov. 3, 2023); see also Doe v. Napa Valley Unified Sch. Dist., No. 17-cv-03753, 2018 WL 4859978, at *2 (N.D. Cal. Apr. 24, 2018) (rejecting defendants' attempt “to reframe Plaintiff's negligence claim as three separate claims, and then move to dismiss two of Plaintiff's three theories of negligence”).¹⁰

While not required based on the preceding ruling, the Court nevertheless addresses a component of the “directed to children” issue given it impacts whether content from third-party sites may be considered in evaluating a COPPA violation. The Court does so beginning with the statute.

Specifically, COPPA provides:

(a) Acts prohibited

(1) In general

It is unlawful for an operator of a website or online service directed to children, or any operator that has actual knowledge that it is collecting personal information from a child, to collect personal information from a child in a manner that violates the regulations prescribed under subsection (b).

15 U.S.C. § 6502(a). The referenced “regulations prescribed” are set forth in 16 C.F.R., Subch. C, Pt. 312, also known as the COPPA Rule. The FTC's COPPA Rule regurgitates the statutory language then imposes a discrete set of required procedures, i.e., a set of means by which a qualifying operator can violate COPPA:

General requirements. ․ Generally, under this part, an operator must:

(a) Provide notice on the Web site or online service of what information it collects from children, how it uses such information, and its disclosure practices for such information (§ 312.4(b));

(b) Obtain verifiable parental consent prior to any collection, use, and/or disclosure of personal information from children (§ 312.5);

(c) Provide a reasonable means for a parent to review the personal information collected from a child and to refuse to permit its further use or maintenance (§ 312.6);

(d) Not condition a child's participation in a game, the offering of a prize, or another activity on the child disclosing more personal information than is reasonably necessary to participate in such activity (§ 312.7); and

(e) Establish and maintain reasonable procedures to protect the confidentiality, security, and integrity of personal information collected from children (§ 312.8).

16 C.F.R. § 312.3.¹¹ Further, pursuant to 15 U.S.C. §§ 6503 and 6505, a violation of the COPPA Rule “shall be treated as a violation of a rule defining an unfair or deceptive act or practice under section 18(a)(1)(B) of the [FTC] Act (15 U.S.C. § 57a(a)(1)(B)).” 16 C.F.R. § 312.9; New Mexico ex rel. Torrez v. Meta Platforms, Inc., No. 23-cv-01115, 2024 WL 413609, at *4 (D.N.M. Feb. 5, 2024).¹²

As to whether a platform is “directed to children,” the COPPA Rule again parrots the Act's definition but elaborates further with a multi-factor test. 16 C.F.R. § 312.2 (definitions). Importantly, in determining whether a website or online service is “directed to children,” the Rule provides for consideration of a platform's:

subject matter, visual content, use of animated characters or child-oriented activities and incentives, music or other audio content, age of models, presence of child celebrities or celebrities who appeal to children, language or other characteristics of the Web site or online service, as well as whether advertising promoting or appearing on the Web site or online service is directed to children. The Commission will also consider competent and reliable empirical evidence regarding audience composition, and evidence regarding the intended audience.

16 C.F.R. § 312.2. The FTC has explained this test “look[s] to the totality of the circumstances” such that “no single factor will predominate over another.” Amendments to the Children's Online Privacy Protection Rule, 78 Fed. Reg. 3972, 3984 (Jan. 17, 2013).

Little credible debate exists that the legal structure requires a fact-intensive analysis to analyze the “directed to children” test. The Court will not engage in such inquiries on a motion to dismiss.

The real issue concerns third parties. Assuming the truth of the States' allegations, a vast majority of the content on Meta's platforms rendering the platforms “child-directed” is third-party content. Meta urges that this third-party content should not convert a general audience service into a “child-directed” one under COPPA. In support, Meta argues (i) 15 U.S.C. § 6501(10)(B) and COPPA FAQ D.2 ¹³ support excluding third-party material from consideration and (ii) interpreting COPPA to impose liability based on third-party content is inconsistent with Section 230 of the Communications Decency Act (“CDA”).

As to the first, subsection (10)(B) simply states that a platform will not be considered “directed to children solely for referring or linking to a commercial website or online service directed to children by using information location tools.” 15 U.S.C. § 6501(10)(B). Referring or linking to a website is materially different from hosting that third-party content. Meta's hyper-technical reading of subsection (10)(B) of Section 6501 does not persuade. As to the second regarding COPPA FAQ D.2, the Court disagrees. Neither the plain language of the statute, nor the promulgated rule compel such a blanket interpretation. The FAQ does not control. Neither the statute nor the Rule make any distinction between platform-created and third-party content, and the Court discerns no basis to impose one.¹⁴

As to statutory consistency, Meta cites generically to Get Oil Out! Inc. v. Exxon Corp., 586 F.2d 726, 729 (9th Cir. 1978) (In construing the Deepwater Port Act, the court held it is “our obligation to so construe federal statutes so that they are consistent with each other, as by this means congressional intent can be given its fullest expression.”). Meta does not argue that assessment of third-party content under COPPA is prohibited by Section 230, only that such assessment is inconsistent with the spirit of Section 230. As to any proposed inconsistency, Meta has not shown that COPPA and Section 230 cannot coexist.¹⁵ Courts have recognized that each statute has the protection of children as a goal,¹⁶ and that they do so in different ways does not render them inconsistent with each other.¹⁷

The Court HOLDS that third-party content hosted by a platform can be considered when determining whether the platform, or a portion thereof, is “directed to children” under COPPA.

IV. SECTION 230

A. Overview

Meta asserts that Section 230 of the CDA “largely bars” the States' consumer protection claims.

As outlined earlier, the States' consumer protection claims fall into two buckets: “deceptive acts and practices” and “unfair and/or unconscionable acts and practices.” (See supra Section I.C.) The first focus on Meta's misrepresentations (affirmative and by omission); the latter focus on the harm caused by specific product features (eight identified in this motion). Each set of claims presents unique theories requiring distinct treatment under Section 230.

As a general proposition, theories and claims differ. The Multistate Complaint does not explicitly identify a “failure-to-warn claim.” The 217-page complaint, in fact, makes use of the terms “warn” or “warning” on only five occasions. (Multistate Compl. ¶¶ 187, 333, 349, 379, 790.) That said, the complaint is explicit in that it seeks to impose liability on Meta's alleged “fail[ure] to disclose the dangerous nature of its Social Media Platforms and because Meta utilized psychologically manipulative engagement-inducing features, knowing that young users are especially susceptible to those psychologically manipulative tactics.” (E.g., id. ¶ 935 (Count XIII: Unfair Acts or Practices by Meta in Violation of Georgia Fair Business Practices Act).)¹⁸ While not clear, the Court finds it sufficient and provides guidance herein so as not to delay these proceedings with yet another amendment. The Court understands that the “deceptions” and “omissions” encompassed by the States' deceptive acts and practices claims include failure-to-warn theories.

B. Legal Framework

The States do not dispute that Section 230 applies to state consumer protection laws and claims asserted by state attorneys general. See, e.g., Zango, Inc. v. Kaspersky Lab, Inc., 568 F.3d 1169, 1177 (9th Cir. 2009) (recognizing that Section 230 “provide[s] immunity from state unfair competition and false advertising actions” (citing Perfect 10, Inc. v. CCBill LLC, 488 F.3d 1102, 1118–19 (9th Cir. 2007))); Free Speech Coal., Inc. v. Colmenero, No. 23-cv-917, 2023 WL 5655712, at *26 (W.D. Tex. Aug. 31, 2023) (“The text of the CDA is clear: ‘No cause of action may be brought and no liability may be imposed under any State or local law that is inconsistent with this section.’ 47 U.S.C. § 230(e)(3). ‘[A]ny’ state law necessarily includes those brought by state governments ․” (alteration in original)), rev'd and vacated in part, 2024 WL 982225, at *17–19 (5th Cir. Mar. 7, 2024) (determining Section 230 did not conflict with or preempt the at-issue Texas statute).¹⁹

Subsection 230(c)(1) provides that “[n]o provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.” 47 U.S.C. § 230(c)(1). The Ninth Circuit has framed this language, alongside subsection 230(e)(3),²⁰ as providing a three-part test to determine whether a defendant is entitled to Section 230 immunity:

[Subsection 230](c)(1) only protects from liability (1) a provider or user of an interactive computer service (2) whom a plaintiff seeks to treat, under a state law cause of action, as a publisher or speaker (3) of information provided by another information content provider.

Barnes v. Yahoo!, Inc., 570 F.3d 1096, 1100–01 (9th Cir. 2009) (footnote omitted).

Prong 1. Meta points out that the parties have never disputed that Meta is “a provider ․ of an interactive computer service” by way of its Facebook and Instagram platforms for the purposes of these claims. Social Media I, 702 F. Supp. 3d at 825. Remaining for elaboration, as in the Court's prior order, are the “second and third factor[s]” which “tend to overlap in significant ways.” Id. at 827 (quoting In re Apple Inc. App Store Simulated Casino-Style Games Litig., 625 F. Supp. 3d 971, 978 (N.D. Cal. 2022)).

Prong 2. As described in the Court's prior order, the second prong of a Section 230 analysis focuses on “whether ‘the duty the plaintiff alleges’ stems ‘from the defendant's status or conduct as a publisher or speaker.’ ” Lemmon v. Snap, Inc., 995 F.3d 1085, 1091 (9th Cir. 2021) (quoting Barnes, 570 F.3d at 1107). A claim meets this prong where the claim is based on “behavior that is identical to publishing or speaking.” Barnes, 570 F.3d at 1107 (emphasis supplied). Critically, Section 230 does not create immunity simply because publication of third-party content is relevant to or a but-for cause of the plaintiff's harm. The issue is whether the defendant's alleged duty to the plaintiff could “have been satisfied without changes to the content posted by the website's users and without conducting a detailed investigation.” Doe v. Internet Brands, Inc., 824 F.3d 846, 851 (9th Cir. 2016).

With respect to the term “publishing” itself, courts understand it to mean “deciding whether to publish or to withdraw from publication third-party content.” Id. The most basic example of online publishing subsection 230(c)(1) is intended to protect is a message board on which content is posted by third parties. For example, Dyroff v. Ultimate Software Grp., Inc., 934 F.3d 1093 (9th Cir. 2019) held the plaintiff could not proceed on a claim that the defendant, an online platform, contributed to the death of her son who purchased heroin from another user on the platform through message board postings. In this circumstance, the defendant was a publisher of third-party content; a congressionally protected status.

“Publishing” also includes editorial decisions and functions ancillary to the decision to make content available. Thus, publishing has been found to “involve[ ] reviewing [and] editing,” such as “review[ing] material submitted for publication, perhaps edit[ing] it for style or technical fluency,” Barnes, 570 F.3d at 1102, and “deciding whether to exclude material,” Fair Hous. Council of San Fernando Valley v. Roommates.Com, LLC, 521 F.3d 1157, 1170–71 (9th Cir. 2008). In general, it is any conduct “rooted in the common sense and common definition of what a publisher does.” Barnes, 570 F.3d at 1102 (also “deciding whether to publish, withdraw, postpone or alter content” and other of “ ‘publisher's traditional editorial functions’ ”) (quoting Zeran v. Am. Online, Inc., 129 F.3d 327, 331 (4th Cir.1997)). The Ninth Circuit has also indicated that Section 230 may provide immunity for any claim that would “necessarily require an internet company to monitor third-party content.” HomeAway.com, Inc. v. City of Santa Monica, 918 F.3d 676, 682 (9th Cir. 2019); Lemmon, 995 F.3d at 1092.

Prong 3. The third prong overlaps with the prior two and concerns information provided by another information content provider. Social Media I, 702 F. Supp. 3d at 827. No further articulation is required.

* * *

In interpreting the second and third prongs, the Court in its prior order focused on the analysis in Fair Hous. Council of San Fernando Valley v. Roommates.Com, LLC, 521 F.3d 1157, 1164 (9th Cir. 2008). Roommates established a test for determining if a platform's actions in altering or presenting content constitutes development, namely, whether the platform provides “neutral tools” for the creation or dissemination of content, which does not destroy Section 230 immunity. Roommates.Com, 521 F.3d at 1172. However, if the platform's conduct materially alters the content, enhancing its alleged illegality, the tool is not neutral, and Section 230 does not bar liability. Id. at 1174–75 (“Where it is very clear that the website directly participates in developing the alleged illegality ․ immunity will be lost.”).

C. Unfair and Unconscionable Acts: Platform Development and Design

Here, Meta moves to dismiss under Section 230 the States' claims based on alleged unfair or unconscionable business practices that target the product development and design of specific features of Meta's platforms.

These theories are subject to the same conduct-specific, feature-by-feature analysis that the Court undertook in its prior order relating to products liability claims.²¹ The Court divides the analysis between those features for which it holds Section 230 provides immunity versus those where it does not.

1. Features Previously Barred by Section 230

With respect to those features where the Court already found they were protected under Section 230, the common thread identified showed that the features “directly target defendants' roles as publishers of third-party content.” Social Media I, 702 F. Supp. 3d at 830. The five are, as identified by the Court:

• Infinite scroll and autoplay features (Multistate Complaint ¶ 847(b));

• Ephemeral content features (id. ¶ 847(c));

• Disruptive audiovisual and vibration notifications and alerts (id. ¶ 847(c));

• The quantification and display of “Likes” (id. ¶ 847(b)); and

• Meta's algorithmic service of content according to “variable reinforcement schedules” (id. ¶ 847(d)).

With respect to these, the States nonetheless argue that the design, development, and deployment of each constitutes an unfair or unconscionable act or practice. The Court addresses each feature:

Infinite scroll and autoplay. The States challenge the “infinite scroll” and “autoplay” features of Meta's platforms. (Multistate Compl. ¶¶ 847(b)–(c); see id. ¶¶ 377, 381; see also id. ¶ 273; Florida Compl. ¶¶ 28, 32.) The Court has held that addressing this alleged product defect “would necessarily require defendants to publish less third-party content.” Social Media I, 702 F. Supp. 3d at 831. That ruling likewise applies to the States' allegations.

Ephemeral content. The States challenge how “Meta designed, developed, and deployed ․ ephemeral content features” as to “exploit[ ] young users' psychological vulnerabilities.” (Multistate Compl. ¶ 847(c); see id. ¶¶ 387–88; Florida Compl. ¶ 35.) The Court has held that “[e]ditorial decisions such as determining the length of content published and how long to publish content are ‘traditional editorial functions’ immune under Section 230, where exercised with regard to third-party content.” Social Media I, 702 F. Supp. 3d at 832. That ruling applies here.

Notifications. The States challenge similarly how “Meta designed, developed, and deployed disruptive audiovisual and vibration notifications and alerts ․ in a way that ․ exploited young users' psychological vulnerabilities.” (Multistate Compl. ¶ 847(c)–(b); Florida Compl. ¶ 115 (“disruptive alerts”).) The Court has held that “where notifications are made to alert users to third-party content, Section 230 bars plaintiffs' product defect claims.” Social Media I, 702 F. Supp. 3d at 833.²² That ruling applies here.

Likes. The States challenge the “quantification and display of ‘Likes’ ” (Multistate Compl. ¶ 847(b)), which “are a quick way for users to express validation or approval of other users' photos or videos, by clicking or tapping a heart icon or the iconic thumbs-up icon” (id. ¶ 227; see also Florida Compl. ¶¶ 33–34). The Court has held that Section 230 bars “product defect claims” regarding “notifications ․ made to alert users to third-party content,” which “includes notifications that someone has ․ liked a user's post.” Social Media I, 702 F. Supp. 3d at 833. That ruling applies here.

The States argue that the Court has not held as to whether Meta's design choice to display a calculated total of “Likes” on a user's post is barred by Section 230. Meta argues the distinction between notifying a user of a “Like” and the manner in which it is displayed on the post is immaterial for purposes of Section 230. The States respond that their claim relating to Likes, especially in light of Project Daisy (discussed supra Section I.B.2), is more nuanced because they allege that Meta knew its design choice to display total “Likes” on a user's post contributed to negative social comparison. Thus, the claim rests not just on design decisions to tabulate a total “Like” counts, but also to obscure options for users to proactively hide those counts, and so should survive Section 230.

The Court disagrees, finding Meta's authority persuasive. In Kimzey v. Yelp! Inc., the Ninth Circuit characterized Yelp's star-rating system, “which is based on rating inputs from third parties and which reduces this information into a single, aggregate metric,” as the kind of “neutral tool” described in Roommates.Com that operates on user-generated “voluntary inputs” and does not amount to “development.” 836 F.3d 1263, 1270 (9th Cir. 2016); see also Klayman v. Zuckerberg, 753 F.3d 1354, 1358 (D.C. Cir. 2014) (“[A] website does not create or develop content when it merely provides a neutral means by which third parties can post information of their own independent choosing online.”). The aggregation of “Likes” is a content-neutral means of displaying a form of third-party content on Meta's platforms. The States' allegations regarding “Likes” do not survive Section 230.²³

Algorithms. The States challenge how “Meta algorithmically served content to young users, according to ‘variable reinforcement schedules.’ ” (Multistate Compl. ¶ 847(d); id. ¶¶ 151–225.)²⁴ The Court has held that Section 230 immunizes defendants from challenges to their “use of algorithms to determine whether, when, and to whom to publish third-party content” because “[w]hether done by an algorithm or an editor, these are traditional editorial functions that are essential to publishing.” Social Media I, 702 F. Supp. 3d at 833. The States argue, in essence, that their unique allegations merit a different approach.

Courts have consistently dismissed claims under Section 230 alleging harms caused by content recommendation algorithms in general, and Facebook's recommendation algorithms in particular. See Dyroff v. Ultimate Software Grp., Inc., 934 F.3d 1093, 1096 (9th Cir. 2019); Force v. Facebook, Inc., 934 F.3d 53, 66 (2d Cir. 2019) (disagreeing with “contention that Facebook's use of algorithms renders it a non-publisher”); M.P. v. Meta Platforms, Inc., No. 22-cv-3830, 2023 WL 5984294, at *1, 3–4 (D.S.C. Sept. 14, 2023) (rejecting argument that “algorithms and internal architecture of social media sites ․ are well beyond the function of traditional publishers” and concluding Section 230 barred allegations against Facebook's algorithms). Recommendation algorithms, absent allegations to the contrary, serve a “content-neutral” function for the dissemination of third-party content immune from liability under Section 230. See Dyroff v. Ultimate Software Grp., Inc., 934 F.3d 1093, 1096 (9th Cir. 2019); id. at 1098 (content recommendation features, “including algorithms,” were part of website's role “as a publisher of others' content”).

Based on the current landscape of the law, the Court disagrees that the States' allegations merit different treatment. First, while the Multistate Complaint contains allegations specific to Meta's knowledge of this design's harmful effects (e.g., internal analyses of “Negative Appearance Comparison” or “NAC” content as amplified by Meta's recommendation algorithms, see, e.g., id. ¶¶ 195–204), the States' theory of harm caused by Meta's recommendation algorithms is not meaningfully distinct from the personal injury plaintiffs' theory previously reviewed by the Court for the purposes of Section 230. As alleged, the algorithms (i) sequence the presentation of material, (ii) individually curate content for each user, and (iii) amplify content to which each user responds most strongly.

Second, the States' argument that because the algorithms cause harm irrespective of the content recommended only serves to confirm that the algorithms are, in fact, “content-neutral.” See Dyroff, 934 F.3d at 1096. In this vein, the States argue that Meta could “take reasonable measures to design” a safer platform “without altering the content that [its] users generate.” Lemmon, 995 F.3d at 1092. However, the States have not adequately explained what reasonable modifications to Meta's algorithms would not involve changing how Meta decides “whether, when, and to whom to publish third-party content.” Social Media I, 702 F. Supp. 3d at 833. Allegations related Meta's recommendation algorithms are barred by Section 230.²⁵

2. Features Previously Not Barred by Section 230

The Court previously held that allegations relating to other forms of disputed features are not barred by Section 230.²⁶ “The defects pled as part of such allegations are not equivalent to speaking or publishing and can be fixed by defendants without altering the publishing of third-party content.” Social Media I, 702 F. Supp. 3d at 829. Here, these are:

• Appearance-altering filters (Multistate Complaint ¶¶ 333–68);

• Features relating to restricting time spent on the platform (id. ¶¶ 578–90); and

• Instagram's “multiple accounts” function (id. ¶¶ 370–72).

As above, the Court considers the arguments presented that each feature and its design, development, and deployment constitutes an unfair or unconscionable act or practice.

Appearance-altering filters. The States challenge Meta's promotion of platform features that promote or encourage eating disorders and body dysmorphia in youth, despite Meta's representations to the contrary and its knowledge of expert warnings about the resulting harms. (Id. ¶ 333.)²⁷ The Court previously held that plaintiffs “plausibly allege that the filters are harmful regardless of whether children eventually post the images that they filtered. Plaintiffs allege that children are harmed simply by creating and then seeing their own altered images. No posting or publication is necessary. There is a defect and a harm separate and apart from publication of any third-party content.” Social Media I, 702 F. Supp. 3d at 830. That reasoning applies here.

Hindering time restrictions. The States challenge Meta's “Daily Limit” feature, which Meta allegedly “claimed would enable users to restrict the amount of time they spend on Instagram each day.” (Multistate Compl. ¶ 578.)²⁸ The Court previously held that allegations of Meta's failure to “provid[e] options to users to self-restrict time used on a platform” were not barred by Section 230. Social Media I, 702 F. Supp. 3d at 829. That ruling applies here.

Instagram's “multiple accounts” function. The States challenge Instagram's “multiple accounts” function, “which allows users to register up to five accounts without having to log out of any one account to access another.” (Multistate Compl. ¶ 370 (emphasis supplied).) The Court has not considered this feature.²⁹

The Court recognizes the few cases which have held that “decisions regarding the ‘structure and operation of [a] website’—such as ‘permitt[ing] users to register under multiple screen names’ ”—reflect, in part, “ ‘choices about what content can appear on the website and in what form’ and thus ‘fall within the purview of traditional publisher functions.’ ” Fields v. Twitter, Inc., 200 F. Supp. 3d 964, 972 (N.D. Cal. 2016) (alterations in original) (quoting Jane Doe No. 1 v. Backpage.com, LLC, 817 F.3d 12, 20–21 (1st Cir. 2016)); see also Doe v. Friendfinder Network, Inc., 540 F. Supp. 2d 288, 295 (D.N.H. 2008) (discussing liability for, among others, “a feature that allowed a single individual to post under multiple screen names”).

However, in the context of adolescents, this feature—which is not exclusively focused on content or publishing activity—can serve as an end-run around other features such as those related to time restrictions. Thus, given the procedural posture, at this juncture the Court is unwilling to make a global decision on this issue.

Allegations regarding Instagram's “multiple accounts” function are not necessarily barred by Section 230.

* * *

In sum, Meta's motion to dismiss is GRANTED to the extent the States' allegations of unfair and unconscionable acts and practices target the following platform features: infinite scroll and autoplay; ephemeral content; how Meta designed and deployed audiovisual and vibration notifications and alerts; the quantification and display of likes; and how Meta algorithmically served content to young users. Meta's motion to dismiss is DENIED as to the States' allegations of unfair and unconscionable acts and practices concerning appearance altering filters; platform features that hindered time restrictions; and Instagram's “multiple accounts” function.

D. Deceptive Acts and Practices: Failure-to-Warn Theory

As a threshold issue, the Court notes that Meta does not move to dismiss the States' deceptive acts and practices claims to the extent they are based on affirmative misrepresentations.³⁰ For example, allegations that Meta knowingly misrepresented facts in public reports (like the CSER) about the harm experienced by users on and the addictive quality of its platforms do not implicate Section 230 and proceed to an assessment on the merits. Any theories of deceptive acts and practices that target Meta's alleged affirmative misrepresentations independent of platform features survive Section 230. See Barnes v. Yahoo!, Inc., 570 F.3d 1096, 1107 (9th Cir. 2009) (The “outwardly manifested intention to create an expectation on the part of another ․ generates a legal duty distinct from the conduct at hand, be it the conduct of a publisher ․”).

Rather, on this issue, the States allege that Meta's “constellation of acts which, viewed both individually and holistically, constitute an unfair trade act or practice,” including when Meta failed to speak about this overall effect. Meta urges that Section 230 prohibits this claim.

With respect to the alleged omissions, a tension exists and the law is still developing. Under Section 230, the concern is always whether the defendant is treated “ ‘as a publisher or speaker’ ‘of information provided by another information content provider.’ ” Internet Brands, Inc., 824 F.3d at 850 (quoting Barnes, 570 F.3d at 1102). Where “at root” a “failure to warn claim faults [a platform] for not mitigating, in some way, the harmful effects of” negative user-generated content, the claim ultimately rests on the “deletion, change, or suppression” of activity protected by Section 230. Est. of Bride by & through Bride v. Yolo Techs., Inc., 112 F.4th 1168, 1180 (9th Cir. 2024) (holding Section 230 precluded products liability claims which “asserted that YOLO's app is inherently dangerous because of its anonymous nature, and that previous high-profile suicides and the history of cyberbullying should have put YOLO on notice that its product was unduly dangerous to teenagers”); see also Herrick v. Grindr, LLC, 306 F. Supp. 3d 579, 590 (S.D.N.Y. 2018) (finding that allegations a platform “failed to incorporate adequate protections against” harmful content “is just another way of asserting that [the platform] is liable because it fails to police and remove [harmful] content”), aff'd, 765 F. App'x 586 (2d Cir. 2019).³¹ That said, where an alleged duty to warn arises from the defendant's knowledge of potential harm acquired independently of its publishing activity, imposing that duty to warn would not treat the defendant as a publisher or speaker. See Internet Brands, Inc., 824 F.3d at 851 (“Jane Doe attempts to hold Internet Brands liable for failing to warn her about information it obtained from an outside source about how third parties targeted and lured victims through Model Mayhem.”).³²

In analyzing that tension, Bride, Herrick, and Internet Brands have provided historic instruction. In Bride, plaintiffs sued YOLO Technologies, which had developed a Snapchat extension permitting users to ask public questions and receive anonymous responses. Bride, 112 F.4th at 1172. Plaintiffs “assert[ed] that YOLO's app is inherently dangerous because of its anonymous nature, and that previous high-profile suicides and the history of cyberbullying should have put YOLO on notice that its product was unduly dangerous to teenagers.” Id. at 1179. Among other claims, plaintiffs contended that YOLO's failure to disclose these risks rendered YOLO liable under a failure-to-warn theory of products liability. Id. The Ninth Circuit held that “[a]t root,” this theory “attempt[s] to hold YOLO responsible for users' speech or YOLO's decision to publish it․ [T]he failure to warn claim faults YOLO for not mitigating, in some way, the harmful effects of the harassing and bullying content. This is essentially faulting YOLO for not moderating content in some way, whether through deletion, change, or suppression.” Id. at 1180.

In Herrick, the plaintiff sued a platform called Grindr on the grounds that it lacked built-in safety features. There, plaintiff alleged that his former boyfriend engaged in “a campaign of malicious catfishing” by impersonating plaintiff on the Grindr platform and encouraging “potential suitors” to go to his “home or workplace for sex.” Herrick, 306 F. Supp. 3d at 584. Plaintiff claimed Grindr misled him into believing it could prohibit impersonating profiles and wrongfully refused to locate and remove the at-issue impersonating profiles. Id. Plaintiff sought a remedy including a “warning,” but the court held that the warning would only become necessary if Grindr (as publisher) did “not police or remove objectionable content.” Because the claim was “inextricably related” with respect to Grindr's role with content, Section 230 provided immunity. Id. at 588, 591.³³

By contrast, in Internet Brands, the platform was held liable for failing to warn the plaintiff about external information revealing how “third parties [had] targeted and lured victims through Model Mayhem,” the defendant's platform. Internet Brands, Inc., 824 F.3d at 851. There, plaintiff Jane Doe, an aspiring model, posted information about herself on “Model Mayhem.” Id. at 847. “Unbeknownst to Jane Doe, two persons, Lavont Flanders and Emerson Callum [(the “perpetrators”)], were using Model Mayhem to identify targets for a rape scheme ․” Id. The perpetrators would browse the profiles posted by models on the website, “contact[ ] potential victims with fake identities posing as talent scouts, and lure[ ] the victims to south Florida for modeling auditions,” at which point the perpetrators would drug the victim, rape her, and record the activity for sale as pornography. Id. The perpetrators did not post their own profiles; they used posted information. Id. Jane Doe alleged that Internet Brands had previously learned that the perpetrators “had been criminally charged,” including the “details of the scheme.” Id. at 849.

The court held that Section 230 did not bar Jane Doe's claim because the “failure to warn claim has nothing to do with Internet Brands' efforts, or lack thereof, to edit, monitor, or remove user generated content.” Id. at 852. The duty to warn under California law arose from Internet Brands' possession of this information and did not require the company to alter any content on its website or change how it publishes content. See id.

This Court must determine ultimately, “at root,” whether the States' failure-to-warn theory seeks to treat Meta as a publisher of third-party content. See Bride, 112 F.4th at 1180. At a minimum, and consistent with the Court's conduct-specific approach, alleged failures to warn about known risks of harms attendant to platform features that do not implicate Section 230 as described under the unfair and unconscionable acts prong of the States' consumer-protection claims survive.

Further, among all the litigation against social media companies proceeding in parallel to this MDL, not one court has yet to hold that Section 230 bars such failure-to-warn theories in general:

• Social Media Cases, No. 22STCV21355, 2023 WL 6847378, at *46 (Cal. Super. Ct. Oct. 13, 2023) (“Holding Meta responsible for a failure to warn of the potentially harmful effects of such design features likewise does not fall within the scope of Section 230 immunity. Insofar as Plaintiffs contend that Meta fraudulently concealed from users Meta's research regarding addiction to its products, its data showing that ‘high time spent users’ are disproportionately young, its awareness that teens report Instagram as a source of increased anxiety and depression, and other research findings of potential harm that do not pertain to harm from specific third-party content, warnings Plaintiffs contend should have been provided do not seek to hold Meta liable for third-party content. Meta is not protected from tort liability for its own failure to warn because these adverse effects that allegedly should have been disclosed result from Meta'[s] own conduct, not from any particular content displayed. Meta could have fulfilled its duty to warn of these potential harms without referencing or deleting any content—the duty springs from its capacity as a creator of features designed to maximize engagement for minors, not from its role as publisher.”).

• State v. TikTok Inc., No. 12CV-23-65, at *8–9 (Ark. Cir. Ct. May 15, 2024) (“[T]he State's claims relating to the TikTok app's addictiveness are not barred by Section 230. Defendants represent to consumers that the TikTok app is safe for minors. Yet, Defendants know from their own studies that the app's addictive nature harms minors. Here again, the State is not attempting to hold Defendants liable for the content on the platform. The State's allegations establish that Section 230 of the Communications Decency Act does not bar the State's claims.”);

• State v. Meta Platforms, Inc., No. D-101-cv-2023-02838 (N.M. 1st Jud. Dist. Ct. June 21, 2024) (“[T]he Court understands there are considerable issues related to the difference between design and other non-publisher activities and publisher activities, such as decisions about content, curation, and the like․ [T]he Court DENIES Meta's Motion.”);

• Utah Div. of Consumer Protection v. Meta Platforms, Inc., No. 230908060, at 11 (Utah Dist. Ct. July 25, 2024) (“[I]n a light most favorable to the Division, the Division's claims do not seek to treat Meta as a publisher or speaker. Although Meta does disseminate third-party content, the Division's claims rest on the Defendants' own acts: their use of features and practices to target children into spending excessive amounts of time and their misrepresentations [and omissions] about the safety of those platforms.”);

• State v. Meta Platforms, Inc., No. 23-cv-4453, at 12–13 (Vt. Sup. Ct. July 28, 2024) (“The State alleges that Meta has failed to disclose to consumers its own internal research and findings about Instagram's harms to youth, including ‘compulsive and excessive platform use.’ The alleged failure to warn is not ‘inextricably linked to [Meta's] alleged failure to edit, monitor, or remove [ ] offensive content.’ ” (citation omitted) (alterations in original) (second quotation from Herrick, 764 F. App'x at 591));

• District of Columbia v. Meta Platforms, Inc., No. 2023-CAB-6550, at *24 (D.C. Super. Ct. Sept. 9, 2024) (“If the claim seeks to hold Meta liable for omissions that make its statements about eating disorders misleading, then, as with the omissions regarding the prevalence of harmful third-party content on Meta's platforms, the claim seeks to hold Meta liable for its own false, incomplete, and otherwise misleading representations, not for its publication of any particular third-party content. If the claim seeks to hold Meta liable for breaching a duty to disclose the harms of its platforms' features, including the plastic surgery filter, then the claim is based on Meta's own conduct, not on any third-party content published on its platforms.”).34

Here, the States present their failure-to-warn theories in broad strokes. Recall that the States allege, in particular, that Meta failed to disclose that its platforms are (i) “not psychologically or physically harmful for young users and are not designed to induce young users' compulsive and extended use, when they are in fact so designed,” and (ii) “less addictive and/or less likely to result in psychological and physical harm for young users than its Social Media Platforms are in reality.” (Multistate Compl. ¶ 846(a), (b).) As evident from the Court's prior order on Section 230 and as described above, these alleged psychological and physical harms to young users resulting from platform use can plausibly result from a combination of platform features, some of which are insulated by Section 230 under products liability and unfairness or unconscionability theories, and some of which are not.

Moreover, the approach to assessing Section 230 has been pushed into new territory by at least one federal court of appeals.³⁵ As to the balance, and given the new and myriad opinions of other jurisdictions, the Court declines to dismiss the claims at this early juncture. Thus, because these novel theories test the boundaries of an area of law in some flux, the Court declines to foreclose the States' (and personal injury plaintiffs') failure-to-warn theories as to known risks of addiction attendant to any platform features or platform construction in general as barred by Section 230. The Court's skepticism of these claims is noted, but for now, the claims proceed. Cf. Utah Div. of Consumer Protection, No. 230908060, at 12 (“In a motion for summary judgment, where the facts are in front of the Court, these features might qualify for Section 230 immunity.”).

V. CONSUMER PROTECTION CLAIMS AGAINST META

A. Deceptive Acts and Practices

Thirty states claim Meta has committed deceptive acts and practices under their consumer protection laws. Meta moves to dismiss these claims on four grounds, namely, the alleged representations are: (i) nonactionable, generic statements of safety resembling puffery or opinion, (ii) not sufficiently specific under Rule 9(b), (iii) not material, and (iv) protected under the Noerr-Pennington doctrine. The Court will consider each basis in turn.³⁶

As a preliminary matter, the Court notes that as long as the States have alleged one actionable misrepresentation, the deceptive acts and practices claims survive. Further, each statement must be considered in the context of the alleged deceptive scheme as a whole, and so the Court is unwilling to parse and isolate every alleged misrepresentation as Meta requests.³⁷ See Williams v. Gerber Prod. Co., 552 F.3d 934, 939 n.3 (9th Cir. 2008) (statement that snack was “ ‘nutritious,’ were it standing on its own, could arguably constitute puffery” but declining to dismiss statement as puffery because the “statement certainly contributes, however, to the deceptive context of the packaging as a whole”).

1. Puffery and Opinion

According to Meta, the States' main theory of deception relates to Meta's statements that its services were “safe” for young users. Any statements regarding “safety” or “well-being,” Meta argues, are statements of opinion, not fact, and so are nonactionable.

“[W]hether a statement is deemed puffery is a question of fact to be resolved by the finder of fact except in the unusual case where the answer is so clear that it may be decided as a matter of law.” Commonwealth by Shapiro v. Golden Gate Nat'l Senior Care LLC, 194 A.3d 1010, 1024 (Pa. 2018); In re Chrysler-Dodge-Jeep Ecodiesel Mktg., Sales Pracs., and Prod. Liab. Litig., 295 F. Supp. 3d 927, 1004–05 (N.D. Cal. 2018) (“[W]hether a statement is deemed puffery is generally considered a question of fact, not law, although there are instances in which the answer is so clear that the issue may be decided as a matter of law ․”). To be actionable, “a statement must be ‘specific and measurable’ and capable of being proven true or false.” Azoulai v. BMW of N. Am. LLC, 2017 WL 1354781, at *8 (N.D. Cal. Apr. 13, 2017) (quoting Southland Sod Farms v. Stover Seed Co., 108 F.3d 1135, 1145 (9th Cir. 1997)) (“[T]here is nothing ‘specific and measurable’ about the word ‘safely.’ ”). Courts routinely dismiss misrepresentation claims based on generalized allegations of “safety.” Id. at 27 (citing, e.g., Morris v. Princess Cruises, Inc., 236 F.3d 1061, 1068 (9th Cir. 2001) (representation that consumers “would be safely and adequately served” failed to state a claim because the statement “is devoid of any meaningful specificity”)).³⁸

For example, in In re Yahoo! Inc. Customer Data Sec. Breach Litig., plaintiffs alleged two misrepresentations: (i) that protecting Yahoo!'s systems and user information is “paramount” to ensuring a secure user experience and maintaining user trust, and (ii) that defendants had physical, electronic, and procedural safeguards that comply with federal regulations to protect user information. No. 16-md-02752, 2017 WL 3727318, at *26 (N.D. Cal. Aug. 30, 2017). The court held that the first statement was “mere puffery on which a reasonable consumer could not rely” while the second statement made a “specific, non-subjective guarantee” that merited reasonable reliance. Id.

However, puffery can be actionable when the defendant knows facts that makes the at-issue statement deceitful or makes the statement in a context that renders it deceitful. See, e.g., Jones v. Corus Bankshares, Inc., 701 F. Supp. 2d 1014, 1027–28 (N.D. Ill. 2010) (“[S]tatements that would otherwise amount to puffery can be actionable if the speaker is aware that the statement is deceptive.”); In re Takata Airbag Prod. Liab. Litig., 462 F. Supp. 3d 1304, 1318 (S.D. Fla. 2020) (marketing of safety features can “cross the line from mere puffery to active misrepresentation” when the speaker has “actual knowledge of the alleged safety defect” (first quoting In re Gen. Motors LLC Ignition Switch Litig., 257 F. Supp. 3d 372, 457 (S.D.N.Y. 2017)); Williams, 552 F.3d at 939 n.3 (9th Cir. 2008) (alleged statement resembling puffery “certainly contributes, however, to the deceptive context of the packaging as a whole”).

The alleged misrepresentations in the Multistate Complaint run the gamut of specificity. On one end of the spectrum, the complaint notes that “Instagram's website characterized the Instagram app as a ‘safe and supportive community’ for its users.” (Multistate Compl. ¶ 130.) In isolation, characterizing Instagram as “safe and supportive” hardly resembles a “specific and measurable” statement of fact. See Azoulai, 2017 WL 1354781, at *8.

Some allegations, though, may be considered deceptive in context. Consider the allegation that Adam Mosseri, the head of Instagram, in June 2019 “told CBS in an interview that teen well-being is a top priority. And two years later, in May 2021, Mosseri minimized Instagram's negative impact on teens, characterizing it to reporters as ‘quite small,’ as reported by the Wall Street Journal that September.” (Multistate Compl. ¶ 126.) Yet internal documentation from Meta in March 2019, for example, is alleged to conclude that the “ ‘average net effect’ of Meta's Platforms on its users is ‘slightly negative,’ ” while acknowledging that “social media can be helpful for youth well-being when the use is ‘moderate.’ ” (Id. ¶ 423.) The Court cannot say as a matter of law that Mosseri's representation of Instagram's negative impact as “quite small” is nonactionable when some of Meta's own researchers are alleged to have determined that its average net impact on all users was negative. See Jones, 701 F. Supp. 2d at 1027–28.

Finally, other allegations, despite Meta's urging to the contrary, do resemble direct misrepresentations. Recall Project Daisy, the internal Meta research experiment to determine whether hiding Like counts on Instagram decreases negative social comparison. “[I]n an interview with press in May 2021 about why Meta did not implement Daisy as a default setting, Mosseri stated ‘there was very little impact and the result was neutral,’ and, therefore, Meta made Daisy an opt-in feature.” (Multistate Compl. ¶ 287.) Yet, Project Daisy is alleged to have reduced the negative impact of seeing posts with many likes (id. ¶ 242), and so, contrary to Mosseri's statement, it did not have a “neutral” result, at least as alleged.³⁹

These are but a few examples of the statements alleged in the Multistate Complaint. Altogether, these isolated representations are part of a cohesive whole which, as alleged, form a deceptive scheme by Meta to obfuscate the risks of serious harm stemming from platform use—in other words, a scheme to hide a variety of objective risks to user safety and well-being. Certainly, some of the alleged statements in isolation may be nonactionable, but given the plausible breadth of allegations, the Court cannot say that Meta's statements all constitute nonactionable puffery at this stage. This case does not resemble the situation in Yahoo!, which involved essentially only two alleged misrepresentations. The States here allege a yearslong course of conduct replete with numerous public statements regarding user safety varying widely in character and degree of deceit.

The motion on this ground is denied.

2. Sufficiency of Pleading Falsity under Rule 9(b)

Meta argues that the States' deceptive practices claims “sound in fraud” and so “must state with particularity the circumstances constituting fraud or mistake.” Fed. R. Civ. P. 9(b).⁴⁰ Meta here focuses on three theories of deception presented by the States: misrepresentations as to (i) Meta's under-13 account policy, (ii) the prevalence of harmful content, and (iii) Project Daisy.

Rule 9(b) requires claims that “sound in fraud” plead with particularity the “time, place, and specific content of the false representations as well as the identities of the parties to the misrepresentations.” Swartz v. KPMG LLP, 476 F.3d 756, 764 (9th Cir. 2007) (quoting Edwards v. Marin Park, Inc., 356 F.3d 1058, 1066 (9th Cir. 2004)). A “claim is said to be ‘grounded in fraud’ or to ‘sound in fraud’ ” when the plaintiff “allege[s] a unified course of fraudulent conduct and rel[ies] entirely on that course of conduct as the basis of a claim.” Vess v. Ciba-Geigy Corp. USA, 317 F.3d 1097, 1103 (9th Cir. 2003). Meta argues the plaintiff “must set forth an explanation as to why the statement or omission complained of was false or misleading.” In re GlenFed, Inc. Sec. Litig., 42 F.3d 1541, 1548 (9th Cir. 1994).⁴¹ However, Meta relies too heavily on securities cases rather than common-law fraud cases.

Rule 9(b) requires that a pleading be specific, not that a pleading be free of potential factual disputes. By and large, the states' allegations more than adequately point to the time, place, content, and speaker of the misrepresentations. (See, e.g., Multistate Compl. ¶ 427 (“On December 8, 2021, Mosseri told Congress, ‘I don't believe that research suggests that our products are addictive.’ ”).) Meta's arguments otherwise largely dispute whether certain sets of statements should be characterized as deceptive—typically a factual issue.

Under-13 Account Policy. The States allege that “[d]espite its public-facing claims that users under the age of 13 are not allowed on Instagram, ․ Meta's private internal documents reveal that Meta has coveted and pursued the under-13 Instagram user demographic for years.” (Multistate Compl. ¶ 648.) Even if true, Meta argues these two facts are not contradictory and thus not deceptive—it can be true that Meta both removes under-13 users once detected and is aware of under-13 platform usage.

Meta misapprehends the States' theory of deception. According to the States, Meta's statements were rendered deceptive by pervasive inaction and noncompliance with its under-13 policy. In support, the states plead (i) specific instances of Meta's inaction (e.g., Meta employee “couldn't tell for sure” if user was under 13 despite confirmation from the mother of a 12-year-old girl with four accounts, id. ¶ 668); (ii) how, “[a]s a matter of policy, Meta employees generally do not take action if the reported account does not contain a user bio or photos” (id. ¶ 686); and (iii) internal admissions that Meta does “very little to keep U13s off [the] platform” (id. ¶ 647), among other allegations.

The States have set forth a plausible theory of deception. Selective application of Meta's under-13 policy combined with an alleged pattern of pervasive inaction can render deceptive Meta's reassurances to the contrary.

Prevalence of Harmful Content. The States allege that statements regarding “the incidence or prevalence of negative or harmful user experiences on” Instagram and Facebook were deceptive. Id. ¶ 846(c). The States argue, in particular, that the CSER, or Community Standard Enforcement Reports, which “describe the percentage of content posted on Instagram and Facebook that Meta removes for violating Instagram and Facebook's Community Standards or Guidelines” (id. ¶ 461), are deceptive because Meta publishes only favorable data in its CSER reports and omits relevant information from its TRIPS and BEEF reports. For example, while Meta's CSER report from the third quarter of 2021 “concluded that only ‘0.05-0.06%’ of views on Instagram were of content that violated Meta's standards on bullying and harassment,” contemporaneous internal user survey data indicated that 28.3% of surveyed Instagram users “witnessed bullying on the platform within the last seven days.” (Id. ¶¶ 490–91.)

Again, Meta argues the facts, the death knell in a motion to dismiss. The Court cannot say that, as a matter of pleading, the States have failed to allege actionable deception.

Project Daisy. As already discussed, Project Daisy was a proposal to remove “Like” counts on posts from a user's default view. In public statements, Meta had explained that Project Daisy was not implemented because it was “beneficial for some, and annoying to others,” and that it “ultimately didn't have as strong an impact as we'd hoped,” among other statements. (Multistate Compl. ¶¶ 284–87.) Meta again seeks to argue the alleged falsity of these statements.

As before, the States have plausibly alleged that Meta's representations regarding Project Daisy could be considered deceptive to a finder of fact, particularly if evaluated as one component of a broader pattern. Meta's explanation of Project Daisy's demise need not be “false” to be nonetheless actionable and misleading. As alleged, Meta researchers “HUGELY pushed” for Project Daisy as a “default for teens.” (Id. ¶¶ 261–62; id. ¶ 290 (Project Daisy “is one of the clearest things (supported by research) that we can do to positively impact social comparison and well-being on” Instagram).) Nonetheless, Project Daisy was projected to damage revenue (id. ¶ 250) and got “stuck in a political war” among Meta's leadership (id. ¶ 264).

The motion on this ground is denied.

3. Materiality of Alleged Deceptions

A representation is considered “material” if the representation would affect a reasonable person's choice of or conduct regarding a product. See, e.g., Kwikset Corp. v. Superior Ct., 51 Cal. 4th 310, 332 (Cal. 2011).⁴² “[T]he materiality of a misrepresentation is typically an issue of fact, and therefore should not be decided at the motion to dismiss stage.” Hinojos v. Kohl's Corp., 718 F.3d 1098, 1107 n.7 (9th Cir. 2013).

With respect to the materiality argument, Meta focuses on statements regarding its alleged goals around increasing users' time spent on its platforms—e.g., Mark Zuckerberg's March 25, 2021, congressional testimony that “it is a common misconception that our teams—our goals, or even have goals [sic], of trying to increase the amount of time that people spend.” (Multistate Compl. ¶ 137.) Other statements are marked as immaterial in Meta's appendix. (See Dkt. No. 517-1.)

Given the Court's assessment above that numerous allegations suffice Rule 9(b)'s pleading standard and constitute actionable misrepresentations as alleged, the Court holds that the States have alleged material misrepresentations that survive a motion to dismiss. The Court declines to isolate and carve out the actionability of individual statements (like the one quoted in the paragraph prior) for the same reasons discussed earlier. The States are permitted to have a factual assessment of individual statements in the context of the alleged scheme of deception as a whole. See Williams, 552 F.3d at 939 n.3.

The motion on this ground is denied.

4. Noerr-Pennington

Meta claims that all of its executives' statements to Congress are nonactionable because congressional testimony is protected under the First Amendment following the Noerr-Pennington doctrine.⁴³

To determine whether a defendant's conduct is immunized under Noerr-Pennington, courts consider “(1) whether the lawsuit imposes a burden on petitioning rights, (2) whether the alleged activities constitute protected petitioning activity, and (3) whether the statute at issue may be construed to avoid that burden.” B&G Foods N. Am., Inc. v. Embry, 29 F.4th 527, 535 (9th Cir.) (cleaned up), cert. denied, 143 S. Ct. 212 (2022).

Meta appears to argue that Noerr-Pennington immunizes any false statements made in legislative proceedings. That reading is unsupportable. Noerr-Pennington immunizes only “protected petitioning activity.” B&G Foods, 29 F.4th at 535. Because Meta has presented “no evidence that [it] was seeking any redress from Congress that implicates its First Amendment right to petition,” and “[i]n light of the underdeveloped factual record and the abbreviated legal arguments,” the Court declines to dismiss any aspect of the States' claims under Noerr-Pennington at this stage. See In re Apple Inc. Sec. Litig., No. 19-cv-02033-YGR, 2020 WL 2857397, at *18 (N.D. Cal. June 2, 2020).

The motion on this ground is denied.

B. Unfair and/or Unconscionable Business Acts or Practices

Meta moves to dismiss consumer protection claims under the unfair and unconscionable acts or practices prong asserted by twenty-seven of the thirty-four states. The Court has held that Section 230 bars the States' unfairness claims relating to many platform features, except for allegations relating to appearance-altering filters, features that hinder time restrictions, and the “multiple accounts” function. These claims proceed to an assessment on the merits.

With respect to the balance, the parties first dispute the standard applied by each state for unfair and/or unconscionable acts or practices prong of the consumer protection claims. More particularly, the applicability of Section 5 of the Federal Trade Commission (“FTC”) Act, the Sperry factors,⁴⁴ or other state-specific standards. Because at least one state puts every test at issue, the Court considers each.

1. Section 5 of the FTC Act

Under Section 5 of the FTC Act, “an unfair practice or act is one that ‘[1] causes or is likely to cause substantial injury to consumers [2] which is not reasonably avoidable by consumers themselves and [3] not outweighed by countervailing benefits to consumers or to competition.’ ” FTC v. Neovi, Inc., 604 F.3d 1150, 1155 (9th Cir. 2010) (quoting 15 U.S.C. § 45(n)). Meta contends that the States' allegations fail under all three elements.

Substantial Injury. “In most cases ‘substantial injury’ involves monetary harm.” FTC v. Direct Mktg. Concepts, Inc., 569 F. Supp. 2d 285, 299 (D. Mass. 2008), aff'd, 624 F.3d 1 (1st Cir. 2010); Am. Fin. Servs. Ass'n v. FTC, 767 F.2d 957, 972 (D.C. Cir. 1985) (“In elaborating the term ‘substantial injury’ in its Policy Statement, the Commission stated that in most cases substantial injury would involve monetary harm ․”). Even where not monetary, “[a]n act or practice can cause ‘substantial injury’ by doing a ‘small harm to a large number of people, or if it raises a significant risk of concrete harm.’ ” Neovi, 604 F.3d at 1157 (quoting Am. Fin. Servs. Ass'n, 767 F.2d at 972); see also FTC v. Accusearch, Inc., No. 06-cv-105, 2007 WL 4356786, at *8 (D. Wyo. Sept. 28, 2007) (noting in addition to economic harm evidence of “a host of emotional harms that are substantial and real and cannot fairly be classified as either trivial or speculative” caused by the business practice), aff'd, 570 F.3d 1187 (10th Cir. 2009).

The States plausibly allege that Meta's purported unfair acts and practices cause substantial injury to consumers, even if that injury is not monetary. First, Meta's acts and practices—as alleged—present a significant risk of “concrete” injuries. For instance, body image and eating disorders are real conditions that can be diagnosed by medical professionals. See Neovi, 604 F.3d at 1157. Second, knowingly developing tools that encourage youth addiction “cannot fairly be classified as either trivial or speculative,” Accusearch, 2007 WL 4356786 at *8, even if the Court accepted Meta's implicit premise that addiction is a nontangible, emotional harm. Third, the States' allegations present a substantial risk of imposing at least a “small harm to a large number of people,” see Neovi, 604 F.3d at 1157, given these practices are allegedly targeted at all minor users of Facebook and Instagram. This argument fails.

Not Reasonably Avoidable. “In determining whether consumers' injuries were reasonably avoidable, courts look to whether the consumers had a free and informed choice.” Neovi, 604 F.3d at 1158. Meta relies primarily on three cases, two of which concern gambling—Phillips, Ristic, and Davis—and none of which concern adolescents.

In Phillips v. Double Down Interactive LLC, the plaintiff alleged that she played Double Down Casino's games through her Facebook account, wagering and losing over $1,000 worth of chips over the course of two years. 173 F. Supp. 3d 731, 736 (N.D. Ill. 2016). In alleging a claim under the Illinois Consumer Fraud and Deceptive Practices Act's unfairness prong, she argued that Double Down's games “are designed to encourage illegal gambling and to exploit the psychological triggers associated with gambling and addiction.” Id. at 743. Because the court found that plaintiff “could have picked other forms of entertainment,” it held Double Down's games could reasonably have been avoided and did not constitute unfair conduct. Id.⁴⁵

Similarly, in Ristic v. Machine Zone, Inc., the defendant produced a videogame called “Game of War,” which included an in-game feature known as the “Casino.” No. 15-cv-8996, 2016 WL 4987943, at *1 (N.D. Ill. Sept. 19, 2016). “Casino” was a virtual slot machine that required real-dollar wagers (via purchased “Chips”) to receive in-game prizes. Id. Plaintiff, who spent over $500 on “Chips” for “Casino,” alleged that defendant engaged in unfair conduct by violating Illinois' gambling laws. Id. at *1–2. Here plaintiff pled that his injuries could not reasonably have been avoided given defendant's “use of psychological triggers that prompt regular consumers, as well as addicts, to continue wagering.” The court disagreed, holding that the plaintiff “made a voluntary choice ․” Id. at *4.

In Davis v. HSBC Bank Nevada, N.A., the plaintiff claimed that defendant's advertisements for a Reward Zone MasterCard were unfair because they lacked disclosure of an annual fee. 691 F.3d 1152, 1158, 1167 (9th Cir. 2012). While not disputed, the advertisement contained the disclaimer that “other restrictions may apply.” Id. at 1168–69. This, combined with a “boldface and oversized” notice to review the terms and conditions, gave plaintiff ample notice of possible additional fees and an opportunity to abort the application. Id. at 1169. Further, the allegedly unfair annual fee “was also avoidable after the account was opened.” Id.

All of Meta's cases are distinguishable. At this preliminary stage the Court cannot say that Meta's allegedly unfair business practices are reasonably avoidable as a matter of law. While Phillips is comparable insofar as the court held that a product designed to be addictive was reasonably avoidable, the case did not address the unique susceptibility of minors, and whether gambling minors would have such a “free and informed choice” as to unknown risks of addiction. Neovi, 604 F.3d at 1158.⁴⁶ Ristic is distinguishable on similar grounds, and Davis is largely inapposite given that Meta does not claim it provided notice of addictive risks.

This reasoning applies with equal force to a review of individual platform features.⁴⁷ As to Meta's “Daily Limit” feature, Meta argues that the states concede the feature is reasonably avoidable because they allege the “feature was designed so that the user can easily dismiss the notification” indicating they have reached their maximum time spent “and return to using Instagram unimpeded.” (Multistate Compl. ¶ 579.) The States respond that this feature must be considered in the context of Meta's broader scheme to design its platform to encourage compulsive use by minors. While the Court has dismissed some allegations involving compulsive use as to specific features under Section 230, the Court cannot say as a matter of law that a minor still retains that “free and informed choice” with respect to use of the “Daily Limit” feature where the minor has developed—by the platform's own design—a compulsive habit to click through and continue use. In addition, the “multiple accounts” function could serve as an end run around the Daily Limits.

Meta also argues that the States' COPPA allegations fail under the “not reasonably avoidable” prong of Section 5 because a child's misrepresentation as to their age is a deliberate choice. In so arguing, Meta effectively advocates for an exception to its “actual knowledge” of under-13 platform users because those children were dishonest. This argument fails. Under COPPA, the statute requires a website operator with “actual knowledge” it is collecting personal information from a child to comply with the COPPA Rule. 15 U.S.C. § 6502(a)(1). Meta has conceded actual knowledge as alleged (only for the purposes of its 12(b)(6) motion), and so it must comply with the COPPA Rule. The Court will not construe the plain language differently.⁴⁸

Not Outweighed by Countervailing Benefits. With respect to the last element, the States must plead that the alleged consumer injuries were “not outweighed by countervailing benefits to consumers or to competition.” Neovi, 604 F.3d at 1158; Parziale v. HP, Inc., 445 F. Supp. 3d 435, 446 (N.D. Cal. 2020) (dismissing claim where plaintiff “fail[ed] to raise any allegation that his injury is not outweighed by a countervailing benefit to consumers”). Given the national outrage, Meta's arguments here strain credulity.

The States identify allegations of the known range of serious harms resulting from compulsive use of Instagram and Facebook, and specifically the negative “net effect” of platform use on young users. See Multistate Compl. ¶ 119 (“[D]ue to Meta's deliberate design choices, the net result has been that young users are negatively affected by using Instagram compulsively.”); id. ¶ 423 (“[An internal] document concludes the ‘average net effect’ of Meta's Platforms on its users is ‘slightly negative.’ ”); id. ¶ 613 (discussing internal email in which Meta employee states “there is increasing scientific evidence (particularly in the US ․ ) that the average net effect of [Facebook] on people's well-being is slightly negative”).

Meta calls these allegations “conclusory.” They are not. The States adequately plead that the alleged harm to Meta's users was not outweighed by any countervailing benefits to its users.⁴⁹

Meta's arguments under this section fail as noted.

2. The Sperry Factors

Some jurisdictions rely on the factors set out in FTC v. Sperry & Hutchinson Co., to determine whether a practice is “unfair.” 405 U.S. at 244 n.5.

a) Whether Sperry Applies

Meta asserts there are nineteen states that adopt or otherwise incorporate the FTC Act's approach to determining unfair acts and practices. (Dkt. No. 517 at 39–41.) The States argue that this oversimplifies variations among state laws which are often less restrictive than the FTC Act, following instead, for example, the “cigarette rule,” also known as the Sperry factors test. (Dkt. No. 599 at 35–36.) Given the Court's ruling as to Section 5, this issue is largely academic as even those claims survive.

Nonetheless, the Court notes that some states permit their courts to consider decisions interpreting Section 5 of the FTC Act as persuasive authority when defining unfair acts and practices. See, e.g., 815 Ill. Comp. Stat. Ann. 505/2 (“In construing this section consideration shall be given to the interpretations of the [FTC] and the federal courts relating to Section 5(a) of the [FTC] Act.”). A review of Illinois's caselaw, for example, indicates the nuanced interplay between such states' consumer-protection laws and Section 5 of the FTC Act:

In determining whether particular conduct is unfair, the [Illinois Consumer Fraud Act (“ICFA”)] dictates that “consideration shall be given to the interpretations of the Federal Trade Commission (FTC) and the federal courts relating to Section 5(a) of the Federal Trade Commission Act.” 815 ILCS 505/2; see Robinson, 775 N.E.2d at 960. The most important of those interpretations are the “Sperry factors,” which the FTC has published and the Supreme Court has cited with approval. Id. See Federal Trade Comm'n v. Sperry & Hutchinson Co., 405 U.S. 233, 92 S.Ct. 898, 31 L.Ed.2d 170 (1972). Illinois recognizes the Sperry test, and so we too will use it as our point of departure.

Batson v. Live Nation Ent., Inc., 746 F.3d 827, 830 (7th Cir. 2014); see also Kahn v. Walmart Inc., 2024 WL 3282097, at *9 (7th Cir. July 3, 2024) (applying Sperry under ICFA).

Meta argues that statutory provisions like the above-cited from Illinois indicates that ICFA and its counterparts should be interpreted consistent with Section 5 of the FTC Act. Meta is incorrect. The Court does not read these statutes to indicate, as Meta advocates, that these state legislatures seek for federal law to control their consumer protection laws. Rather, they more likely intended to permit courts to apply their state-specific consumer-protection standards and merely consider “unfairness” under Section 5 as persuasive, rather than binding, authority. This interpretation is bolstered when considering that the Batson court then explicitly turned to assess an ICFA claim under the Sperry factors, not Section 5. Moreover, Meta's lead case, Phillips, applied the Sperry factors under ICFA. See Phillips, 173 F. Supp. 3d at 742 (citing Batson, 746 F.3d at 830). Violations of ICFA are alleged in Count 40 of the Multistate Complaint. (See Multistate Compl. ¶¶ 946–48.) The Sperry test is thus at issue.⁵⁰

b) Applying the Sperry Factors

Under Sperry, a court considers whether a practice (i) offends public policy, (ii) is immoral, unethical, oppressive, or unscrupulous, or (iii) is substantially injurious to consumers. See, e.g., Colgate v. JUUL Labs, Inc., 402 F. Supp. 3d 728, 758 (N.D. Cal. 2019). In some jurisdictions, like Illinois, “all three of the criteria in Sperry do not need to be satisfied to support a finding of unfairness.” Batson, 746 F.3d at 830 (quoting Robinson v. Toyota Motor Credit Corp., 775 N.E.2d 951, 961 (Ill. 2002)). At least in Illinois, the third factor—whether a practice is substantially injurious—can be assessed under the factors laid out in Section 5 of the FTC Act. See Kahn, 2024 WL 3282097, at *10.

Meta rests belated arguments under Sperry on the same points it made under Section 5's purportedly stricter standard, which the Court holds the States have satisfied. For the same reasons, the Court holds the States have alleged unfair acts and practices under the Sperry factors. Meta otherwise has failed to meet its burden for dismissal because it has not advanced any other Sperry-specific arguments.

3. State-Specific Standards

Meta separately moves to dismiss the unfair or unconscionable acts or practices claims of California, Colorado, Indiana, Kansas, New Jersey, and Oregon.⁵¹ The Court considers each.

a) California

Historically, California courts have relied on three different tests for assessing unfair or unconscionable acts or practices. See Davis v. HSBC Bank Nevada, N.A., 691 F.3d 1152, 1169 (9th Cir. 2012). First, prior to Cel-Tech, courts applied the Sperry balancing test that weighs “the utility of the defendants conduct against the gravity of the harm to the alleged victim,” known as the South Bay test in California. Id. at 1169.⁵² Second, the California Supreme Court in Cel-Tech adopted a new test which requires that any finding of unfairness be “tethered” to “some legislatively declared policy or proof of some actual or threatened impact on competition.” Cel-Tech Commc'ns, Inc. v. Los Angeles Cellular Tel. Co., 973 P.2d 527, 544 (Cal. 1999). Third, some courts have “borrowed the three-pronged test set forth in the FTC Act.” Davis, 691 F.3d at 1170. That said, the statute is routinely used. Further, this Court has previously explained:

The UCL prohibits “any unlawful, unfair or fraudulent business act or practice and unfair, deceptive, untrue or misleading advertising.” Cal. Bus. & Prof. Code § 17200. The UCL creates a cause of action for business practices that are (1) unlawful, (2) unfair, or (3) fraudulent. Each “prong” of the UCL provides a separate and distinct theory of liability. Lozano v. AT & T Wireless Servs., Inc., 504 F.3d 718, 731 (9th Cir. 2007)․

․

The unlawful prong of the UCL prohibits “anything that can properly be called a business practice and that at the same time is forbidden by law.” Cel-Tech Commc'ns., Inc. v. L.A. Cellular Telephone Co., 20 Cal.4th 163, 180, 83 Cal.Rptr.2d 548, 973 P.2d 527 (1999) (quotation marks and citations omitted). “By proscribing ‘any unlawful’ business practice, the UCL permits injured [parties] to ‘borrow’ violations of other laws and treat them as unlawful competition that is independently actionable.” In re Adobe Sys., Inc. Privacy Litig., 66 F. Supp. 3d 1197, 1225 (N.D. Cal. 2014) (quoting Cel-Tech Commc'ns., 20 Cal.4th at 180, 83 Cal.Rptr.2d 548, 973 P.2d 527).

Doe v. Epic Games, Inc., 435 F. Supp. 3d 1024, 1049–50 (N.D. Cal. 2020).⁵³

Here, the California legislature has declared in the California Age-Appropriate Design Code Act (“CAADCA”) that “[b]usinesses that develop and provide online services, products, or features that children are likely to access should consider the best interests of children when designing, developing, and providing that online service, product, or feature.” Cal. Civ. Code § 1798.99.29. The legislature intended for CAADCA to “promote innovation by businesses whose online products, services, or features are likely to be accessed by children by ensuring that those online products, services, or features are designed in a manner that recognizes the distinct needs of children at different age ranges.” 2022 Cal. Legis. Serv. Ch. 320, § 1(c) (A.B. 2273) (West). This policy recognized the “bipartisan agreement at the international level, in both the United States and in the State of California, that more needs to be done to create a safer online space for children to learn, explore, and play.” Id. § 1(a)(3).

The States acknowledge, and Meta emphasizes, that CAADCA has preliminarily been held as unconstitutional, at least in part. NetChoice, LLC v. Bonta, No. 22-CV-08861-BLF, 2023 WL 6135551 (N.D. Cal. Sept. 18, 2023) (granting preliminary injunction enjoining enforcement of the CAADCA and holding CAADCA violated the First Amendment), affirmed in part and vacated in part, No. 23-2969, 2024 WL 3838423, at *2 (9th Cir. Aug. 16, 2024).⁵⁴ Courts have prohibited the enforcement of “an unconstitutional statute indirectly by invoking the UCL.” Daghlian v. Devry Univ., Inc., 2007 WL 5625508, at *8 n.40 (C.D. Cal. Dec. 10, 2007). The States argue that “the Legislature's mere failure to prohibit an activity does not prevent a court from finding it unfair.” Cel-Tech, 973 P.2d at 542.

The Court finds that, even if Cel-Tech applied in this action,⁵⁵ California's UCL claims are sufficiently “tethered” to a “legislatively declared policy.” Id. at 544. As specified above, the California legislature has made clear its goal with respect to children and online safety. 2022 Cal. Legis. Serv. Ch. 320, § 1(a)(3), (c). While Bonta made some unconstitutional findings, it specifically did not find unconstitutional the legislature's overarching goal motivating the enactment of CAADCA. See Bonta, 692 F. Supp. 3d at 936–37 (“[A]lthough the stated purpose of the Act—protecting children when they are online—clearly is important, NetChoice has shown that it is likely to succeed on the merits of its argument that the provisions of the CAADCA intended to achieve that purpose do not pass constitutional muster.”). Here, California's UCL claim targets conduct separate and apart from the activity CAADCA sought to regulate. As such, California's UCL claim does not seek to enforce an unconstitutional statute, see Daghlian, 2007 WL 5625508, at *8 n.40, and is fairly characterized as tethered to legislative policy, see Cel-Tech, 973 P.2d at 544. The motion as to California fails.

b) Colorado

Colorado alleges that Meta committed a violation of the Colorado Consumer Protection Act (“CCPA”) under subection 6-1-105(rrr). (Multistate Compl. ¶ 892.) Meta argues that the CCPA does not recognize claims of unfair practices.

To prove a violation of the CCPA, the attorney general must first show that “the defendant engaged in an unfair or deceptive trade practice.” Hall v. Walter, 969 P.2d 224, 235 (Colo. 1998); id. at 236 (distinguishing elements required for claims brought by the attorney general). Section 6-1-105 of the CCPA is titled “Unfair or deceptive trade practices.” Colo. Rev. Stat. Ann. § 6-1-105. Subsection 6-1-105(1) enumerates dozens of means by which a “person engages in a deceptive trade practice.” Id. § 6-1-105(1) (emphasis supplied). That subsection lists as a “deceptive trade practice” when a person “[e]ither knowingly or recklessly engages in any unfair, unconscionable, deceptive, deliberately misleading, false, or fraudulent act or practice.” Id. § 6-1-105(1)(rrr) (emphasis supplied).⁵⁶ Subsection 6-1-105(3), by contrast, clarifies that the “deceptive trade practices listed in this section are in addition to and do not limit the types of unfair trade practices actionable at common law or under other statutes of this state.” Id. § 6-1-105(3) (emphasis supplied). One court has written that the “implication of this language” in subsection 6-1-105(3) “is that the foregoing provisions proscribe only deceptive conduct and that nondeceptive unfair trade practices are actionable only if brought under other statutory or common law causes of action such as the Colorado Antitrust Act.” In re HIV Antitrust Litig., No. 19-cv-02573, 2023 WL 3006572, at *2 (N.D. Cal. Apr. 18, 2023).

Courts confronted with interpreting subsection 6-1-105(rrr) have declined to adopt a construction of the CCPA “that would sweep in anticompetitive conduct.” In re HIV Antitrust Litig., 2023 WL 3006572, at *2; see also Sheet Metal Workers Loc. 441 Health & Welfare Plan v. GlaxoSmithKline, PLC, 737 F. Supp. 2d 380, 408 (E.D. Pa. 2010) (noting “the absence of any case from Colorado allowing an unfair trade practice claim to proceed”); In re New Motor Vehicles Canadian Exp. Antitrust Litig., 350 F. Supp. 2d 160, 180 n.28 (D. Me. 2004) (“Although the word ‘unfair’ appears in these decisions, I have found no Colorado case where the court applied the term substantively to broaden the CCPA's reach to unfair, in addition to deceptive, practices.”).

As the Colorado Supreme Court has recognized, (i) “[t]he CCPA does not list all the industries to which it applies, nor does it specify all the types of transactions it covers,” (ii) the list of “deceptive or unfair practices” “in section 6-1-105 is not exhaustive,” and (iii) limiting “the only practices covered [to] those specifically listed in the act” would “severely hinder[ ] the CCPA's broad remedial power.” Showpiece Homes Corp. v. Assurance Co. of Am., 38 P.3d 47, 54 (Colo. 2001). It then emphasized that “[i]n enacting the statute, the General Assembly could not have possibly enumerated all, or even most, of the practices that the CCPA was intended to cover,” id., and relied on the language in Sperry that “[i]t is impossible to frame definitions which embrace all unfair practices,” id. (quoting Sperry, 405 U.S. at 240).

Thus, Meta's motion to dismiss Colorado's CCPA claims fails on three grounds. First, none of Meta's cases concern an AG's oversight of a business's practices.

Second, even if the Court were to accept Meta's construction of the CCPA and strictly follow the distinction proposed in In re HIV Antitrust Litig. outside of the antitrust context, Meta does not explain why the States' allegations fail to constitute a “deceptive unfair” or a “deceptive unconscionable” act or practice under subsection 6-1-105(1)(rrr). The States allege Meta developed, marketed, and distributed platform features it knew would cause or exacerbate harms to children without any warning to the children or their parents. Meta provides no standard by which to evaluate these allegations under subsection 6-1-105(1)(rrr).⁵⁷ The Court cannot assume Colorado's theory fails.

Third, Meta's proposed limitation of the CCPA cuts too broadly. The Court is not persuaded that the Colorado Supreme Court would read the CCPA to exclude all claims of unfair or unconscionable acts or practices, especially in light of its language in Showpiece. Moreover, Meta's construction would effectively strike from the CCPA its language that covered conduct includes “knowingly or recklessly engag[ing] in any unfair, [or] unconscionable ․ act or practice,” which the legislature added to the CCPA on May 23, 2019. See Colo. Rev. Stat. Ann. § 6-1-105(1)(rrr).⁵⁸ The cases holding to the contrary are focused on anticompetitive conduct which, as explained in In re HIV Antitrust Litig., are covered by other sources like the Colorado Antitrust Act. See 2023 WL 3006572, at *2.

Meta has not sustained its burden as to warrant dismissal of Colorado's claims of unfair and unconscionable acts or practices under the CCPA.

c) Indiana

Indiana alleges that Meta “has engaged in unfair, abusive, and/or deceptive acts, omissions, and/or practices” in violation of Ind. Code Ann. § 24-5-0.5-3(a), (b)(1), and (b)(2). (Multistate Compl. ¶¶ 967–69.) Meta argues that Indiana does not permit standalone unfairness claims. The Court disagrees.

Indiana's Deceptive Consumer Sales Act (“DCSA”) “is a remedial statute and ‘shall be liberally construed and applied to promote its purposes and policies’ of protecting consumers from deceptive or unconscionable sales practices.” Kesling v. Hubler Nissan, Inc., 997 N.E.2d 327, 332 (Ind. 2013) (quoting Ind. Code Ann. § 24-5-0.5-1). Under the DCSA, a “supplier may not commit an unfair, abusive, or deceptive act, omission, or practice in connection with a consumer transaction.” Ind. Code Ann. § 24-5-0.5-3.⁵⁹ The DCSA “does not define ‘unfair,’ ‘abusive,’ or ‘deceptive.’ ” Crum v. SN Servicing Corp., 2021 WL 3514153, at *2 (S.D. Ind. Aug. 10, 2021). “Indiana courts tend to conflate the three words.” Id.

Nonetheless, Indiana courts have recognized claims of unfairness. See, e.g., Gasbi, 120 N.E.3d at 618 (holding allegations that defendant “charged an unfair consumer fee and did not state its intention as part of the bargaining process” fell “generally within the realm of the” DCSA); Bank v. Huizar, 178 N.E.3d 326, 338 (Ind. Ct. App. 2021) (allegations that defendant's “agents' refusal to leave the residence after being asked and locking themselves in the car until the keys were provided also constitute unfair and abusive acts” sufficed for standalone DCSA claim).

Contrary to Meta's assertion, Indiana courts have recognized claims of unfairness, see Gasbi, 120 N.E.3d at 618, including standalone claims of unfairness, see Huizar, 178 N.E.3d at 338.⁶⁰ Indiana's claim under the unfairness prong survives.

d) Kansas

Kansas alleges that Meta has engaged in unconscionable acts and practices in violation of of Kan. Stat. Ann. § 50-627. (Multistate Compl. ¶ 980.) Meta asserts that Kansas's claim fails because it lacks allegations of “deceptive bargaining conduct.”

The Kansas Consumer Protection Act (“KCPA”) prohibits a supplier from engaging in any “deceptive act or practice” or “unconscionable act or practice” in connection with a consumer transaction. Kan. Stat. Ann. §§ 50-626, 50-627. The KCPA “shall be construed liberally to promote,” among other policies, the protection “of consumers from suppliers who commit deceptive and unconscionable practices.” Id. § 50-623. “Unconscionability is not defined in the KCPA,” but the statute sets forth a non-exhaustive set of example unconscionable acts in Section 50-627(b). State ex rel. Stovall v. DVM Enterprises, Inc., 62 P.3d 653, 657 (Kan. 2003).⁶¹ The Supreme Court of Kansas has explained that “the State need[s] to prove some element of deceptive bargaining conduct and unequal bargaining power to prove unconscionability” and that “there is no claim under the KCPA without evidence of any deceptive or oppressive practices, overreaching, intentional misstatements, or concealment of facts.” Stovall, 62 P.3d at 659.

Meta provides no authority explaining what qualifies as “deceptive bargaining conduct.” As set forth above, the States' complaint does present actionable allegations of deceptive conduct which are sufficient to state a claim under the KCPA.

e) New Jersey

New Jersey alleges that Meta has committed “unconscionable or abusive commercial practices in violation of” the New Jersey Consumer Fraud Act (“CFA”). (Multistate Compl. ¶ 1048.)

Meta argues that claims of unconscionability in New Jersey are required to be deceptive and so fail for the same reasons as the deceptive acts and practices claims discussed under Kansas law. See Kugler v. Romain, 279 A.2d 640, 652 (N.J. 1971) (“[I]n consumer goods transactions ․ unconscionability must be equated with the concepts of deception, fraud, false pretense, misrepresentation, concealment and the like.”); Pullman v. Alpha Media Pub., Inc., 2013 WL 1286144, at *5 n.3 (S.D.N.Y. Mar. 28, 2013) (“Plaintiff's allegations of an ‘unconscionable business practice’ fail for the same reason” as her deceptiveness claims.), aff'd, 624 F. App'x 774 (2d Cir. 2015).⁶²

Because Meta's only basis for dismissing New Jersey's unconscionable acts and practices claim is for a failure to allege deception, and the Court has held that the States' deceptive acts and practices claims survive, New Jersey's claims of unconscionability survive.

f) Oregon

Oregon alleges that Meta has “employed unconscionable tactics in violation of” Oregon's Unlawful Trade Practices Act (“UTPA”), Or. Rev. Stat. Ann. § 646.607(1). (Multistate Compl. ¶ 1107.) Meta argues that this claim fails because “unconscionable tactics” have been interpreted to require the offending business to have tricked customers to enter into a transaction.⁶³

The UTPA prohibits “any unconscionable tactic in connection with selling, renting or disposing of real estate, goods or services, or collecting or enforcing an obligation.” Or. Rev. Stat. Ann. § 646.607. The UTPA provides a non-exhaustive set of four forms of unconscionable tactics, which includes when a person (a) “[k]nowingly takes advantage of a customer's physical infirmity, ignorance, illiteracy or inability to understand the language of the agreement;” (b) “[k]nowingly permits a customer to enter into a transaction from which the customer will derive no material benefit;” (c) “[p]ermits a customer to enter into a transaction with knowledge that there is no reasonable probability of payment of the attendant financial obligation in full by the customer when due;” or (d) “[k]nowingly takes advantage of a customer who is a disabled veteran, a disabled servicemember or a servicemember in active service, or the spouse of a disabled veteran, disabled servicemember or servicemember in active service.” Id. § 646.605(9).

The Oregon Supreme Court has relied on “the common-law doctrine of unconscionability” to define the term “unconscionable tactics” and elaborate on the sort of relationship it entails between a plaintiff and defendant. Gordon v. Rosenblum, 393 P.3d 1122, 1128 (Or. 2017). “Although today the doctrine appears most often in contract law, historically the doctrine has been applied in many areas of law where courts sought to avoid what they perceived as unfair outcomes.” Id. Consistent with that flexible application, it “has followed that pattern and has applied the doctrine in a variety of contexts.” Id.⁶⁴ In particular, the court wrote that, with respect to the last example of unconscionable tactics in section 646.605(9), “[t]he notable absence of a transaction or agreement in that example suggests that a transaction or agreement is not a required element of ‘unconscionable tactics.’ ” Id. at 1130.⁶⁵

Gordon makes clear that “unconscionable tactics” is a flexible concept that can encompass theories of unconscionability beyond the examples provided in section 646.605(9). Meta's argument fails for relying on too narrow a construction of the term. Even so, the States' theory of unconscionability is similar to the conduct described in section 646.605(9)(a)—that Meta has “[k]nowingly take[n] advantage of a customer's physical infirmity, ignorance, illiteracy or inability to understand the language of [an] agreement.” Or. Rev. Stat. Ann. § 646.605(9)(a). As discussed repeatedly in this order, the States allege Meta developed, marketed, and distributed platform features it knew would cause or exacerbate harms to children and specifically preyed on children's unique susceptibility to addictive features. The States plausibly allege “unconscionable tactics” under Oregon law.

* * *

Meta's motion to dismiss the States' unfair and/or unconscionable acts and practices claims is DENIED as to the merits of those claims. Plaintiffs' have presented sufficient allegations to withstand Meta's motion to dismiss under Section 5 of the FTC Act, the Sperry factors, and the various state-specific considerations Meta has presented.

C. Statutory Scope of Consumer Protection Claims

Meta seeks dismissal of the States' and personal injury plaintiffs' consumer-protection claims for lack of a connection to commerce under myriad state's statutes. Meta organizes states into three buckets. First, five states require a connection with a “consumer transaction” (Indiana, Kansas, Ohio, Utah, and Virginia) and Maryland with “consumer goods or services.” Second, nineteen states require claims to those in the conduct of any “trade or commerce” (namely, Alabama, Connecticut, Georgia, Hawai'i, Idaho, Illinois, Kentucky, Louisiana, Maine, Massachusetts, Michigan, Missouri, Nebraska, Pennsylvania, Rhode Island, South Carolina, Tennessee, Texas, and Vermont). Third, three states limit private causes of action to “consumers” (Alabama, Texas, and Vermont). The Court considers each set.

1. “Consumer Transaction” or “Consumer Goods or Services”

Meta claims the laws of Indiana, Kansas, Ohio, Utah, and Virginia require a connection to a “consumer transaction” to state a consumer protection claims and Maryland similarly requires a sale of consumer goods or services.⁶⁶

The plain language of the relevant statutes in Indiana, Kansas, Ohio, Utah, and Virginia limit claims to those where practices occur “in connection with a consumer transaction.” See:

• Ind. Code Ann. § 24-5-0.5-3(a) (“A supplier may not commit an unfair, abusive, or deceptive act, omission, or practice in connection with a consumer transaction.”);

• Kan. Stat. Ann. § 50-626(a) (“No supplier shall engage in any deceptive act or practice in connection with a consumer transaction.”);

• Ohio Rev. Code Ann. § 1345.02(A) (“No supplier shall commit an unfair or deceptive act or practice in connection with a consumer transaction.”);

• Utah Code Ann. § 13-11-4(1) (“A deceptive act or practice by a supplier in connection with a consumer transaction violates this chapter ․”);

• Va. Code Ann. § 59.1-200(A) (prohibiting “fraudulent acts or practices committed by a supplier in connection with a consumer transaction”).

Next, “consumer transaction” is defined in unique but similar ways. Thus:

• In Indiana: it includes the “disposition of ․ a service, or an intangible.” Ind. Code Ann. § 24-5-0.5-2(a)(1) (“ ‘Consumer transaction’ means a sale, lease, assignment, award by chance, or other disposition of an item of personal property, real property, a service, or an intangible ․”).

• In Kansas: it includes any “disposition for value of property or services,” Kan. Stat. Ann. § 50-624(c), for which Kansas courts require the plaintiff to “allege facts one could characterize as her exchanging anything of value with the defendants to secure their services.” Berry v. Nat'l Med. Servs., Inc., 205 P.3d 745, 752 (Kan. Ct. App. 2009) (emphasis supplied), aff'd, 257 P.3d 287 (Kan. 2011).

• In Ohio: it includes any “transfer of ․ an intangible,” regardless of whether money changes hands. See Ohio Rev. Code Ann. § 1345.01(A) (“ ‘Consumer transaction’ means a sale, lease, assignment, award by chance, or other transfer of an item of goods, a service, a franchise, or an intangible ․”); Williams v. Edwards, 717 N.E.2d 368, 373 (Ohio Ct. App. 1998) (“[T]he fact that no money changed hands has no bearing on the applicability of the [Ohio Consumer Sales Practices Act] ․”).67

• In Utah: it includes the “disposition of goods” and “services, ․ both tangible and intangible.” Utah Code Ann. § 13-11-3(2)(a) (“ ‘Consumer transaction’ means a sale, lease, assignment, award by chance, or other written or oral transfer or disposition of goods, services, or other property, both tangible and intangible ․”).

• In Virginia: it includes the “advertisement, sale, lease, license or offering for sale, lease or license, of goods or services to be used primarily for personal, family or household purposes.” Va. Code Ann. § 59.1-198.

Maryland's Consumer Protection Act (“MCPA”) does not rely on the term “consumer transaction” but includes, among other interactions, “[t]he sale, lease, rental, loan, or bailment of any consumer goods, consumer realty, or consumer services.” Md. Code Ann., Com. Law § 13-303(1) (emphasis supplied).

Each state requires a liberal construction of its consumer protection law to effectuate its remedial purpose. See:

• Indiana: Ind. Code Ann. § 24-5-0.5-1 (requiring the Act to be “liberally construed” to “modernize the law governing deceptive and unconscionable consumer sales practices” and to “protect consumers”);

• Kansas: Moore v. Bird Eng'g Co., P.A., 41 P.3d 755, 760 (Kan. 2002) (The Act must “be construed liberally to streamline the law of consumer transactions and to protect consumers from unscrupulous suppliers.” (citing Kan. Stat. Ann. § 50-623));

• Maryland: Washington Home Remodelers, Inc. v. State, Off. of Att'y Gen., Consumer Prot. Div., 45 A.3d 208, 219 (Md. 2012) (The Act “constitutes remedial legislation that is intended to be construed liberally in order to promote its purpose of providing a modicum of protection for the State's consumers.”);

• Ohio: Anderson v. Barclay's Cap. Real Est., Inc., 989 N.E.2d 997, 1000 (Ohio 2013) (“The [Act] is remedial in nature, having been designed to compensate for incomplete consumer remedies available at common law,” and courts must “liberally construe the statute in favor of the consumer.”);

• Utah: Fell v. Alco Cap. Grp. LLC, 538 P.3d 1249, 1256 (Utah Ct. App. 2023) (“[T]he [Act] ‘shall be construed liberally to promote,’ among other things, the protection of ‘consumers from suppliers who commit deceptive and unconscionable sales practices.’ ” (quoting Utah Code Ann. § 13-11-2));

• Virginia: Ballagh v. Fauber Enterprises, Inc., 773 S.E.2d 366, 368 (Va. 2015) (The Act “shall be applied as remedial legislation to promote fair and ethical standards of dealing between suppliers and the consuming public” and is construed liberally in favor of the injured party (quoting Va. Code Ann. § 59.1-197)).

The courts of at least two states—in Indiana and Utah—have decided and currently agree that the provision of social media platforms to users without monetary consideration constitutes a “consumer transaction” under each state's consumer-protection law.⁶⁸

The Court of Appeals of Indiana reversed a trial court which had held that downloading the free TikTok app is not a consumer transaction, because it does not involve an exchange for money. State v. TikTok Inc., No. 23A-PL-3110, 2024 WL 4340387, at *8–9 (Ind. Ct. App. Sept. 30, 2024).⁶⁹ As noted above, the Indiana Deceptive Consumer Sales Act (“DCSA”) defines a “consumer transaction” to mean “a sale, lease, assignment, award by chance, or other disposition of an item of personal property, real property, a service, or an intangible ․” Ind. Code Ann. § 24-5-0.5-2(a)(1). The appellate court wrote:

We agree with the State that DCSA's statutory definition of a consumer transaction does not include the words “exchange for money.” There is no question that any of the described dispositions in the statutory definition can be, and we presume most often are, effectuated by way of an exchange for money, but the statutory language does not require such an exchange.

․ [T]he plain and ordinary definition of the word “sale,” which is not otherwise defined in the DCSA, includes any consideration to effectuate the transfer of property, not only an exchange for money.

It is undisputed that TikTok exchanges access to its app's content library for end-user personal data. That is the bargain between TikTok and its end-users. And, under the plain and ordinary use of the word, that is a “sale” of access to TikTok's content library for the end-user's personal data. TikTok's business model is therefore a consumer transaction under the DCSA.

Further, TikTok's argument that we limit consumer transactions to exchanges for money not only disregards the plain and ordinary meaning of the word “sale” but also narrows the scope of the DCSA beyond its plain terms, which is expressly contrary to the DCSA's requirement that we “liberally” interpret its provisions.

State v. TikTok Inc., 2024 WL 4340387, at *9.

In Utah Div. of Consumer Protection v. Meta Platforms, Inc., the Third Judicial District Court of Utah held, under a set of facts similar to those before this Court, that Utah's Division of Consumer Protection (the “Division”) “allege[d] a sale and/or a transfer sufficient to support a consumer transaction under the” Utah Consumer Sales Practices Act. No. 230908060, at 13 (Utah Dist. Ct. July 25, 2024).⁷⁰ The court explained:

The Division alleges facts that could support a conclusion the transaction involves a sale or transfer, including that Meta does not provide its products for free, that its business model allows it to turn attention into money, and that when users spend time on Meta's social media platforms, Meta harvests their personal data (including interests, religion, beliefs, and what they like to watch or buy). The Complaint alleges that Utah consumers' personal data has value to the Defendants as Meta has built an entire business model on enticing users to give up their personal data and their time as consideration for their services on the Defendants' platforms.

Id. at 13–14.

Here, Meta provides access to its platform in exchange for its users' personal data and time spent viewing, among other content, advertisements which generate profit for Meta. This is an exchange for valuable consideration. A social media company cannot exempt itself from consumer-protection law because its consumers exchange their personal data and time for a product or service rather than cash for goods.⁷¹ These states' consumer-protection statutes are not so restrictive, and Meta has failed to persuade otherwise.

The Court agrees with the analysis of both cases discussed and holds that the States and personal injury plaintiffs have adequately alleged consumer-protection claims “in connection with a consumer transaction” under Indiana, Kansas, Ohio, and Virginia law, and likewise for Maryland.

2. “Trade or Commerce”

Meta argues that nineteen states ⁷² limit claims to those involving “trade or commerce,” which Meta contends does not exist here.⁷³

With some minor variation in the language of conduct covered, the Court agrees that each state prohibits unfair or deceptive acts in “the conduct of any trade or commerce.”⁷⁴ Most states define “trade or commerce,” again with little material variation, to “mean[ ] the advertising, the sale or rent or lease, the offering for sale or rent or lease, or the distribution of any services and any property, tangible or intangible, real, personal or mixed, and any other article, commodity, or thing of value in this state.” Conn. Gen. Stat. Ann. § 42-110a(4).⁷⁵

“Trade and commerce,” as these definitions make clear, is in general “broadly defined,” consistent with the remedial character of these states' consumer-protection statutes and typical requirement they be “liberally construed in favor of those whom the legislature intended to benefit.” Fink v. Golenbock, 680 A.2d 1243, 1259 (Conn. 1996). The States are uniform in this view.⁷⁶ The states present various formulations of how to determine whether an activity takes place in “trade or commerce,” focusing on the exchange of valuable consideration and the business purpose behind the activity.⁷⁷

Variations on this theme do not impact the final analysis. Yes, some states do emphasize that their consumer-protection laws “do[ ] not prohibit any deceptive acts or unfair practices that are in some nebulous way connected to trade or commerce.” G & G Closed Cir. Events, LLC v. Castillo, 327 F. Supp. 3d 1119, 1133 (N.D. Ill. 2018) (emphasis supplied); Indian Hills Civic Ass'n v. Indian Lake Prop. Owners Ass'n, Inc., 637 S.W.3d 622, 632 (Mo. Ct. App. 2021) (“ ‘Commerce’ under the MMPA is not so wide as to cover every economic interaction contemplable ․”).⁷⁸ Some states explain that a “violation may not be alleged for activities that are incidental to an entity's primary trade or commerce.” McCann Real Equities Series XXII, LLC v. David McDermott Chevrolet, Inc., 890 A.2d 140, 164 (Conn. Ct. App. 2006). For instance, “transactions resulting not from ‘the conduct of any trade or business’ but rather from ‘private negotiations between two individual parties who have countervailing rights and liabilities established under common law principles of contract, tort and property law' remain beyond the purview of” these statutes. Hoehl Fam. Found., 2020 WL 10790035, at *16 (citation omitted); see also Binette, 688 A.2d at 907 (contrasting transactions made “in a business context” with those “done on a private, nonprofessional basis”). Whether the defendant “was engaged in the conduct of trade or commerce is a fact-intensive query that requires the courts to consider the totality of the circumstances.” Fayne v. Vincent, 301 S.W.3d 162, 174 (Tenn. 2009). This case is far from those.

Interestingly, at least one court has reached the issue of whether language like “trade or commerce” can encompass free use of a social media platform. In State v. Meta Platforms, Inc., the Vermont Superior Court held that “Instagram is certainly engaging in commerce when it provides Instagram to Vermonters,” because “[u]sers of Instagram allegedly sign a contract agreeing to give Instagram access to personal information for advertising purposes in exchange for using the site.” No. 23-cv-4453, at 15 (Vt. Sup. Ct. July 28, 2024). “That sufficiently pleads a contractual business relationship between Instagram and each user.” Id. The court also noted that “Meta's argument that there must be money changing hands is unsupported,” because “[t]here is no requirement of a specific purchase and sale in” cases brought by the state. Id. at 16.

The Court agrees with the Vermont court's approach and holds that the States and personal injury plaintiffs have adequately alleged unfair or deceptive acts or practices “in the conduct of any trade or commerce” under the laws of the challenged states. This alleged exchange of users' use of Meta's platforms for their personal data is Meta's “primary trade or commerce,” at least insofar as its Facebook and Instagram platforms are concerned. See McCann Real Equities Series XXII, LLC, 890 A.2d at 164. At a minimum, Meta cannot credibly argue that providing its platform to users does not qualify as a “distribution of [a] service.” See, e.g., 815 Ill. Comp. Stat. Ann. 505/1(f). Thus, the alleged unfair and deceptive acts or practices occur “in the conduct of” Meta's “trade or commerce,” regardless of whether Meta's users pay for use of Facebook or Instagram.⁷⁹ Meta has otherwise failed to provide authority to support its burden to show otherwise at this stage.

3. “Consumer”

Finally, Meta moves to dismiss on the grounds that the unfair and/or deceptive acts and practices claims brought by personal injury plaintiffs under the laws of Alabama, Texas, and Vermont all fail because each state limits such private causes of action to “consumers.” Plaintiffs do not dispute the statutory references.⁸⁰ Each state defines “consumer” in unique but similar ways. See Ala. Code § 8-19-3(4) (defining “consumer” as “[a]ny natural person who buys goods or services for personal, family, or household use”); Tex. Bus. & Com. Code Ann. § 17.45(4) (“ ‘Consumer’ means an individual ․ who seeks or acquires by purchase or lease, any goods or services ․”); Vt. Stat. Ann. tit. 9, § 2451a (1) (“ ‘Consumer’ means any person who purchases, leases, contracts for, or otherwise agrees to pay consideration for goods or services ․ for the person's use or benefit”).

Thus, the question is whether the relevant personal injury plaintiffs are alleged to have “bought,” “purchased,” or “paid consideration” for use of either Facebook or Instagram. Meta simply advances its say-so interpretation that each of these terms requires the transfer of money. In opposition, plaintiffs note that they “ ‘buy’ access to Defendants' platforms through the provision of valuable personal data and attention.” (Dkt. No. 600 at 21, 28.)

For the first time on reply, Meta points to a line of caselaw from California, not Alabama, Texas, or Vermont, holding that private-plaintiff users of “free” social media applications were not “consumers” under a different statute, California's Consumer Legal Remedies Act (“CLRA”).⁸¹

The Court lacks a substantive response. The terms “bought,” “purchased,” or “paid consideration” often connote the transfer of money. See, e.g., Pay, Black's Law Dictionary (12th ed. 2024) (“1. To give money for a good or service that one buys; to make satisfaction ․” (emphasis supplied)); Purchase, Black's Law Dictionary (12th ed. 2024) (“1. The act or an instance of buying. 2. The acquisition of an interest in real or personal property by sale, ․ or any other voluntary transaction.”); Buy, Black's Law Dictionary (12th ed. 2024) (“buy See PURCHASE (1).”). Further, “Consumer,” as defined by statute, is potentially more restrictive than either “consumer transaction” or “trade or commerce.”

Given the changing economy and no briefing on how the Supreme Courts of Alabama, Texas, and Vermont might approach this question, the Court declines to rule on the issue at this stage with insufficiently tailored argument.

* * *

Meta's motion to dismiss the States' and personal injury plaintiffs' unfair and deceptive acts and practices claims from nineteen states due to the scope of statutory language is DENIED.

D. Restitution

Meta moves to dismiss or otherwise limit the scope of the States' claims for restitution under the consumer-protection laws of twenty-one states.⁸² Meta argues that the particular States have not alleged a loss of money or property restorable to their consumers. The States disagree and, at any rate, argue the recovery available under a theory of restitution is not necessarily so limited.

The availability of restitution as a remedy may be considered on a motion to dismiss. See Sonner v. Premier Nutrition Corp., 971 F.3d 834, 844 (9th Cir. 2020) (affirming district court's dismissal of “claims for equitable restitution under the UCL and CLRA”).

“Employed to denote liability based on unjust enrichment, the word ‘restitution’ is a term of art that has frequently proved confusing.” Restatement (Third) of Restitution and Unjust Enrichment § 1, cmt. c (Am. L. Ins. 2011). “On the one hand, there are significant instances of liability based on unjust enrichment that do not involve the restoration of anything the claimant previously possessed,” notably, cases involving disgorgement of profits. Id. “On the other hand, there are numerous situations in which a claimant's undoubted right to the restitution (or restoration) of something does not depend on the unjust enrichment of the defendant.” Id. Finally, “[c]ategories are further complicated by a set of cases which do involve a literal restitution, and which might logically be explained as a means of avoiding or rectifying the defendant's unjust enrichment, but which, by convention, are not so classified.” Id.

A brief survey of the at-issue states' laws reveals similar variation in how states distinguish or conflate concepts like “restitution,” “disgorgement,” and “unjust enrichment.” Some states' consumer-protection statutes explicitly permit recovery of restitution as more narrowly construed (i.e., restoration of loss) and disgorgement or unjust enrichment. See Ariz. Rev. Stat. Ann. § 44-1528(A)(3) (permitting the court to “[r]equire that any profits, gain, gross receipts or other benefit obtained by means of any practice in this article declared to be unlawful be disgorged and paid to the state for deposit in the consumer remediation subaccount of the consumer restitution and remediation revolving fund”); Colo. Rev. Stat. Ann. § 6-1-110 (“The court may make such orders or judgments as may be necessary to prevent the use or employment by the person of any such deceptive trade practice or that may be necessary to completely compensate or restore to the original position of any person injured by means of any such practice or to prevent any unjust enrichment by any person through the use or employment of any deceptive trade practice.”).

Other states appear to embrace the broader definition of restitution as encompassing disgorgement of ill-gotten gains. See, e.g., State v. LG Elecs., Inc., 340 P.3d 915, 924 n.55 (Wash. Ct. App. 2014) (“Whereas ‘[r]estitution measures the remedy by the defendant's gain and seeks to force disgorgement of that gain,’ damages ‘measures the remedy by the plaintiff's loss and seeks to provide compensation for that loss.’ ” (citation omitted) (alteration in original)), aff'd, 375 P.3d 636 (Wash. 2016).⁸³

Some provide more tailored language which may reflect a narrower conception of restitution—that the court “may make such additional orders or judgments as may be necessary to restore to any person in interest any money or property” acquired by prohibited means. Neb. Rev. Stat. Ann. § 59-1608(2) (emphasis supplied); see also 6 R.I. Gen. Laws Ann. § 6-13.1-5(c). These also provide that “[t]he court may order such additional equitable” or other “relief as it deems necessary to protect the public from further violations.” Neb. Rev. Stat. Ann. § 87-303(a); see also 6 R.I. Gen. Laws Ann. § 6-13.1-5(a) (“[T]he attorney general may bring an action in the name of the state against the person to restrain by temporary or permanent injunction the use of the method, act, or practice, upon the giving of appropriate notice to that person and to seek any other relief that may be appropriate.” (emphasis added)).

At least one state may permit restitution but has been found to exclude disgorgement as an available remedy under its consumer-protection laws. Brown v. Google LLC, No. 20-CV-03664-LHK, 2021 WL 6064009, at *18 (N.D. Cal. Dec. 22, 2021) (“Plaintiffs may not seek disgorgement as a remedy” for California UCL violations.). However, quantifiable non-pecuniary interests, like a “diminution of the value of ․ private and personally identifiable data” may be recoverable as restitution. Id.

Finally, other states present unique statutory frameworks or other considerations as to their attorney general's remedial power that do not fit neatly into the categories described above.⁸⁴

Meta's motion to dismiss asks this Court to limit the restitution available under these states' laws because “[t]he relevant statutes only permit restitution to ‘restore’ any ‘money or property’ ” acquired by unlawful conduct. (Dkt. No. 517 at 54.) As evidenced by the variations in state law described above, the Court disagrees with the broad brush strokes and will not presume that the restitution recoverable under these states' consumer-protection statutes' is so limited. Meta selectively quotes many of these statutes, omitting explicit reference in some statutes to “disgorgement,” “equitable relief,” or other relief “that may be appropriate.” Given these nuances, the Court cannot determine the scope of available “restitution” under these statutes on Meta's say-so interpretation of the statutory language.⁸⁵

Additionally, even under the narrowest construction of restitution—that restitution only encompasses restoration of consumers' loss—the States have suggested a plausible theory of loss to their consumers. As in Brown v. Google, quantifiable non-pecuniary interests, like a “diminution of the value of ․ private and personally identifiable data,” could, in theory, be recoverable as restitution should it be adequately proven. See 2021 WL 6064009, at *18. Given that Meta does not dispute the availability of restitution where explicit in statute, the States' claims for restitution survive regardless of the scope of restitution under any state's consumer protection law because they have presented a viable theory under the narrowest view of restitution.⁸⁶

Meta's motion to dismiss or otherwise limit the states' claims for restitution is DENIED.

VI. PERSONAL INJURY PLAINTIFFS' CLAIMS

Meta, YouTube, Snap, and TikTok seek dismissal of the personal injury plaintiffs' consumer-protection claims, and Meta seeks dismissal of the personal injury plaintiffs' concealment and misrepresentation claims. With respect to plaintiffs' consumer-protection claims against Meta, the Court revisits prior rulings and addresses issues of reliance and ascertainable loss.

A. Personal Injury Plaintiffs' Consumer-Protection Claims Against Meta Only

1. Prior Rulings

In Count 7 of the personal injury plaintiffs' 2AMC, plaintiffs assert unfair and deceptive acts and practices claims under applicable state law.

Prior Rulings. As a threshold issue, the defendants move to dismiss these claims on some of the same grounds already asserted with respect to the States. The Court's reasoning with respect to the States' claims applies to the plaintiffs' consumer-protection claims as well. Thus: (i) allegations based on the design and deployment of certain features are barred by Section 230 (see supra Section IV.C.1); (ii) plaintiffs have adequately pled actionable statements by Meta (see supra Section V.A.1–3); and (iii) defendants have not shown at this stage that its statements to Congress are protected under Noerr-Pennington (see supra Section V.A.4). Further, arguments that the allegations do not concern “trade or commerce” or lack a connection with a “consumer transaction” are denied (see supra Section V.C). The Court has also already addressed Meta's arguments regarding whether certain plaintiffs are “consumers” under relevant consumer protection statutes (see supra Section V.C.3).

Accordingly, the Court DENIES the defendants' motion to dismiss the personal injury plaintiffs' consumer-protection claims to the same extent as assessed above with respect to the States and limits the scope in the same manner.

2. Actual Reliance

Meta argues that plaintiffs have failed to allege actual reliance on the misrepresentations and omissions as, at least, twenty-one states' consumer protection laws require. Plaintiffs dispute the legal requirement for many states, and argue their allegations are sufficient regardless.

a) Disputed State Standards

The parties agree that ten states require a showing of actual reliance, namely, Arizona, California, Georgia, Indiana, Iowa, Nevada, Pennsylvania, South Dakota, Texas, and Virginia. See, e.g., Schellenbach v. GoDaddy.com, LLC, 321 F.R.D. 613, 624 (D. Ariz. 2017) (“Parties that do not actually rely on a false statement or material omission have no claim under the” Arizona Consumer Fraud Act.). The parties dispute the requirement under the remaining eleven states: Colorado, Maine, Massachusetts, Minnesota, New Jersey, North Carolina, North Dakota, Ohio, Oregon, Rhode Island, and South Carolina law. (See Dkt. No. 517 at 60–61 & n.67; Dkt. No. 600 at 30–31.)

With respect to the latter, the issue appears to center on defendants' view that even where a state does not explicitly frame an element in terms of “reliance,” it is effectively required where a litigant must allege a direct causal relationship between an alleged misrepresentation and the plaintiff's harm. The Court disagrees except as to North Carolina.

First, at least three of the eleven states (Minnesota, New Jersey, and Oregon) have explicitly rejected this formulation of causation. Second, defendants' motion ignores plaintiffs' theories of causation, some of which were discussed in the Court's prior order in the context of products liability claims. See Social Media I, 702 F. Supp. 3d at 862 (“[A]n analysis of the ‘common facts’ alleged in plaintiffs' MAC is sufficient for plaintiffs to plausibly allege, at this preliminary stage, that defendants' actions proximately caused plaintiffs' injuries.”). As outlined there, these are sufficient causal allegations. Given defendants' broad approach, the Court declines to dismiss plaintiffs' consumer-protection claims on defendants' view of “reliance,” with the exception of North Carolina, and provides an overview of the disputed states' applicable law, highlighting the portions upon which the decision rests:

Colorado. “[R]eliance does not appear to be an element of a claim under Colorado's consumer protection statute.” Hamilton v. TBC Corp., 328 F.R.D. 359, 381 (C.D. Cal. 2018). “Reliance often provides a key causal link between a consumer's injury and a defendant's deceptive practice.” Garcia v. Medved Chevrolet, Inc., 263 P.3d 92, 98 (Colo. 2011) (en banc) (emphasis supplied).

Maine. “A claim for negligent or intentional misrepresentation requires proof of detrimental reliance upon a material false statement of fact. Similarly, a claim for a deceptive trade practice requires proof of a material misrepresentation that misleads the consumer regarding choice or conduct in relation to a product.” GxG Mgmt., LLC v. Young Bros. & Co., 457 F. Supp. 2d 47, 50 (D. Me. 2006) (citation omitted). Thus, while these standards may serve “similar” purposes, they are distinct as the court's assessment of other states' laws makes clear. Most if not all states require a showing of detrimental reliance for claims of negligent or intentional misrepresentation but do not necessarily so require such a showing for claims of deceptive acts and practices.

Massachusetts. “A successful [Massachusetts General Law, Chapter 93A, § 2] action based on deceptive acts or practices does not require proof that a plaintiff relied on the representation ․” Aspinall v. Philip Morris Companies, Inc., 813 N.E.2d 476, 486 (Mass. 2004); see also id. at 489 (“[T]he plaintiffs [need not] show that each individual consumer relied on the defendants' false promise when purchasing Marlboro Lights.”); Loughlin v. Vi-Jon, LLC, No. 20-cv-11555, 2024 WL 1347092, at *4 (D. Mass. Mar. 29, 2024) (“Plaintiffs need not show proof of actual reliance on a misrepresentation to recover damages under [Chapter] 93A.” (alteration in original) (quoting Casavant v. Norwegian Cruise Line Ltd., 952 N.E.2d 908, 912 (Mass. 2011))).⁸⁷

Minnesota. While the consumer must “prove a ‘causal nexus’ between the plaintiff's injuries and the defendant's wrongful conduct,” “ ‘[a]llegations of reliance are ․ not necessary to state a claim’ ” because “the legislature had eliminated the requirement of pleading and proving ‘traditional common law reliance’ as an element.” Wiegand v. Walser Auto. Groups, Inc., 683 N.W.2d 807, 811 (Minn. 2004) (quoting Grp. Health Plan, Inc. v. Philip Morris Inc., 621 N.W.2d 2, 12 (Minn. 2001)) (alterations in original).

New Jersey. The New Jersey Consumer Fraud Act “does not require proof of reliance” to state a claim. Gennari v. Weichert Co. Realtors, 691 A.2d 350, 366 (N.J. 1997); see also Varacallo v. Massachusetts Mut. Life Ins. Co., 752 A.2d 807, 814 (N.J. App. Div. 2000) (“[C]ommon law fraud requires proof of reliance while consumer fraud requires only proof of a causal nexus between the concealment of the material fact and the loss.”).

North Dakota. In North Dakota, the plaintiff must allege that “the plaintiff has been or is likely to be injured as a result of the false statement ․” N. Bottling Co. v. Henry's Foods, Inc., 474 F. Supp. 3d 1016, 1028 (D.N.D. 2020); see also In re Pharm. Indus. Average Wholesale Price Litig., 252 F.R.D. 83, 98 (D. Mass. 2008) (noting the parties “appear[ed] to agree that” North Dakota's consumer-protection statute was among the state statutes that “do not require the element of reliance” (citing N.D. Cent. Code Ann. § 51-15-01, et seq.)).

Ohio. “Unlike a fraud claim, where a plaintiff must allege harm above and beyond the misrepresentation and reliance thereon, a cause of action accrues under the [Ohio] Consumer Sales Practices Act [(“OCSPA”)] as soon as the allegedly unfair or deceptive transaction occurs.” Delahunt v. Cytodyne Techs., 241 F. Supp. 2d 827, 835 (S.D. Ohio 2003); see also Cicero v. Am. Satellite, Inc., 2011-Ohio-4918, ¶ 9 (Ohio 2011) (“[B]ecause appellant had prior knowledge of the allegedly omitted terms and conditions of the offers, he could not have been deceived by the emails, and, therefore, he was not entitled to recovery under the OCSPA.”).

Oregon. For Oregon's Unlawful Trade Practices Act, “[w]hat a plaintiff must prove is that (1) the defendant committed an unlawful trade practice; (2) plaintiff suffered an ascertainable loss of money or property; and (3) plaintiff's injury (ascertainable loss) was the result of the unlawful trade practice. In other words, plaintiff must suffer a loss of money or property that was caused by the unlawful trade practice. Whether, to prove the requisite causation, a plaintiff must show reliance on the alleged unlawful trade practice depends on the conduct involved and the loss allegedly caused by it.” Pearson v. Philip Morris, Inc., 361 P.3d 3, 28 (Or. 2015); see also Clark v. Eddie Bauer LLC, 532 P.3d 880, 889 (Or. 2023) (discussing Pearson).

Rhode Island. “To adequately plead a claim,” a plaintiff must have (1) been “the subject of a deceptive practice or act in connection with the purchase of a service”; and (2) “suffered an ascertainable loss of money or property as a result of the deceptive practice.” Laccinole v. Assad, 2016 WL 868511, at *7 (D.R.I. Mar. 7, 2016); see also In re Pharm. Indus. Average Wholesale Price Litig., 252 F.R.D. at 98 (noting the parties “appear[ed] to agree that” Rhode Island's consumer-protection statute was among the state statutes that “do not require the element of reliance” (citing 6 R.I. Gen. Laws Ann. § 6-13.1-1, et seq.))

South Carolina. “To bring a successful SCUTPA claim, ‘the plaintiff must show: (1) the defendant engaged in an unfair or deceptive act in the conduct of trade or commerce; (2) the unfair or deceptive act affected public interest; and (3) the plaintiff suffered monetary or property loss as a result of the defendant's unfair or deceptive act(s).’ ” Gault v. Thacher, 367 F. Supp. 3d 469, 485 (D.S.C. 2018) (quoting Wright v. Craft, 640 S.E.2d 486, 498 (S.C. Ct. App. 2006)).

North Carolina. North Carolina, by contrast, appears to require allegations of reliance. “[A] claim under section 75–1.1 stemming from an alleged misrepresentation does indeed require a plaintiff to demonstrate reliance on the misrepresentation in order to show the necessary proximate cause.” In re Hester, 2015 WL 6125308, at *4 (Bankr. E.D.N.C. Oct. 16, 2015) (quoting Bumpers v. Cmty. Bank of N. Virginia, 747 S.E.2d 220, 226 (N.C. 2013)).

The motion as to these states, other than North Carolina, is DENIED.

b) Evaluating Actual Reliance

Where reliance is required, the Court evaluates the relevant allegations, namely for twelve states: Arizona, California, Georgia, Indiana, Iowa, Nevada, North Carolina, Pennsylvania, South Dakota, Texas, and Virginia.

As discussed supra Section IV.B, Rule 9(b) requires claims that “sound in fraud” be pled with particularity. Swartz v. KPMG LLP, 476 F.3d 756, 764 (9th Cir. 2007).⁸⁸ In general, the parties agree that to satisfy Rule 9(b), a plaintiffs must allege that they “read and relied on a specific misrepresentation.” Haskins v. Symantec Corp., 654 F. App'x 338, 339 (9th Cir. 2016). “As for showing reliance on an alleged omission, the standard is more relaxed ․” In re ZF-TRW Airbag Control Units Prod. Liab. Litig., 601 F. Supp. 3d 625, 767 (C.D. Cal. 2022).⁸⁹ “Whether, to prove the requisite causation, a plaintiff must show reliance on the alleged unlawful trade practice depends on the conduct involved and the loss allegedly caused by it.” Pearson v. Philip Morris, Inc., 361 P.3d 3, 28 (Or. 2015).⁹⁰

In short, defendants seek individualized allegations of reliance.⁹¹ Plaintiffs respond that because plaintiffs' allegations are rooted in the defendants' silence, i.e., defendants' material omissions and failures to disclose, individualized allegations of reliance are not necessary. Plaintiffs do target defendants' affirmative statements. (See Dkt. No. 600, Plfs.' Opp. at 7 (“Defendants' misleading statements about their platforms' supposed safety were contrary to concrete information in their possession about harms caused by using their platforms, and were materially misleading in the context of those omissions.”).)⁹²

While plaintiffs can proceed on multiple grounds, where they do so, the elements of each claim must still be satisfied. Here, plaintiffs allege both affirmative misrepresentation and independent actionable omissions on their own. Cf. Pearson v. Philip Morris, Inc., 361 P.3d at 28. Accordingly, plaintiffs' claims of deceptive acts and practices that are based on defendants' affirmative statements fail under the laws of Arizona, California, Georgia, Indiana, Iowa, Nevada, North Carolina, Pennsylvania, South Dakota, Texas, and Virginia.

However, as to claims based on Meta's omissions, plaintiffs have adequately alleged actual reliance under any state's law. As explained supra Section V.A.3 with respect to the States' claims, the information allegedly suppressed by Meta is plausibly material. Moreover, it is plausible that a plaintiff would have become aware of a theoretical disclosure of safety risks in using defendants' platforms, which the plaintiffs allegedly use on a consistent basis.

That dozens of plaintiffs allege they continue to use Meta's services up to the present after the Wall Street Journal published in 2021, is ultimately of no significant import. For pleading purposes, continued use of Meta's services does not exclude the possibility that plaintiffs changed their behavior. Also, disclosure by the Wall Street Journal—whose teen and under-13 readership across the United States the Court does not know—is not the same as disclosure by the social media platform itself. Finally, as the previous two points make clear, this argument veers into disputing the fact of actual reliance, which the Court will not determine on a motion to dismiss. In some ways, this is an academic issue. While claims based on affirmative statements in select states cannot stand absent allegations of reliance, the statements are nonetheless evidence probative of claims based on omissions.

3. Ascertainable Loss

Meta argues that the consumer-protection claims brought by personal injury plaintiffs in at least eighteen states should be dismissed for a failure to allege an ascertainable loss as a result of the alleged unfair or deceptive acts or practices.⁹³

Many states provide a private cause of action under the state's consumer-protection statute to a plaintiff who has suffered, with varying language, “any ascertainable loss of money or property, real or personal, as a result of the ․ act or practice declared unlawful.” E.g., Idaho Code Ann. § 48-608(1).

While the state standards vary as to the requirements at the pleading stage and the substance of what constitutes an “ascertainable loss,” a few general observations merit mention. In general, an ascertainable loss “is a deprivation, detriment [or] injury that is capable of being discovered, observed or established.” Marinos v. Poirot, 66 A.3d 860, 866 (Conn. 2013) (alteration in original) (quoting Serv. Rd. Corp. v. Quinn, 698 A.2d 258, 262 (Conn. 1997)) (discussing ascertainable loss under the Connecticut Unfair Trade Practices Act). Further, the “loss is ascertainable if it is measurable even though the precise amount of the loss is not known,” and “there is no need to allege or prove the amount of the actual loss.” Id. (quoting Serv. Rd. Corp., 698 A.2d at 262). For certain claims, some states look to a “nexus between ascertainable loss and the plaintiff's expected benefit of the bargain.” D'Agostino v. Maldonado, 78 A.3d 527, 541 (N.J. 2013); Furst v. Einstein Moomjy, Inc., 860 A.2d 435, 438 (N.J. 2004) (“ascertainable loss” was “the replacement value of the” customer's defective carpet, “not the purchase price”).

For instance, in some states “economic losses, including medical expenses and lost wages” constitute ascertainable losses. Simms v. Candela, 711 A.2d 778, 781 (Conn. Super. Ct. 1998); see also Little v. Brown & Williamson Tobacco Corp., 1999 WL 33291385, at *11 (D.S.C. Mar. 3, 1999) (finding that “medical costs, lost earnings, and any other alleged out-of-pocket expenses” constitute ascertainable losses). By contrast, other states have held that “allegations of emotional distress” are insufficient to meet the ascertainable loss requirement. Di Teresi v. Stamford Health Sys., Inc., 88 A.3d 1280, 1285 (Conn. Ct. App. 2014). Further, ascertainable loss must be causally related to the defendant's conduct; that is, “[t]he question is whether plaintiffs have adequately pleaded ‘any ascertainable loss’ as ‘a result of’ defendant's alleged misrepresentation ․” Paul v. Providence Health Sys.-Oregon, 240 P.3d 1110, 1120 (Or. Ct. App. 2010), aff'd, 273 P.3d 106, 115 (Or. 2012) (affirming Court of Appeals determination that money “expended to prevent or mitigate” “speculative” future harms “is not the kind of loss compensable under the UTPA”).

In general, personal injury plaintiffs seek recovery for “expenses for hospitalizations and medical treatments, and other economic harm that includes but is not limited to lost earnings and loss of earning capacity,” among other alleged harms. (2AMC at 269.) Allegations of medical expenses, lost earnings, and loss of earning capacity as a result of defendants' alleged unfair and deceptive acts and practices suffice at the pleading stage to show an “ascertainable loss.” The defendants are correct that the extent and character of ascertainable loss will vary by the individual plaintiff—and some plaintiffs may well be unable to show an ascertainable loss—but such a determination is better suited to an analysis of the specific factual allegations brought by a plaintiff under that plaintiff's applicable state law. Defendants are limited, practically speaking, from making such individualized allegations in the context of consolidated motions to dismiss in an MDL with hundreds of plaintiffs. Future dispositive motions with respect to the personal injury bellwether plaintiffs' claims will provide an appropriate ground to precisely challenge evidence of ascertainable loss and the applicable framework.

* * *

Defendants' motion to dismiss Count 7 of the 2AMC is:

GRANTED insofar as plaintiffs seek to impose liability on defendants' affirmative statements under the laws of Arizona, California, Georgia, Indiana, Iowa, Nevada, North Carolina, Pennsylvania, South Dakota, Texas, and Virginia;

DENIED insofar as plaintiffs seek to impose liability on defendants' affirmative statements under the laws of Colorado, Maine, Massachusetts, Minnesota, New Jersey, North Dakota, Ohio, Oregon, Rhode Island, and South Carolina;

DENIED insofar as plaintiffs seek to impose liability on defendants' omissions or failures to disclose under the law of any at-issue state; and

DENIED without prejudice on the issue of ascertainable loss.

B. Personal Injury Plaintiffs' Consumer-Protection Claims Against Snap, YouTube, and TikTok

The YouTube, Snap, and TikTok defendants separately filed a two-page joinder as to Meta's motion to dismiss Count 7 of the 2AMC. (Dkt. No. 519.) As these defendants note, “Rule 9(b) does not allow a complaint to merely lump multiple defendants together.” Swartz v. KPMG LLP, 476 F.3d 756, 764 (9th Cir. 2007). The Court thus considers personal injury plaintiffs' consumer-protection claims against YouTube, Snap, and TikTok separately.

These defendants argue that plaintiffs' consumer-protection claims should fail because plaintiffs fail to present plausible allegations of a commercial transaction by any defendant, any actionable misstatements, or reliance on those misstatements. As elaborated above, the Court finds that plaintiffs have plausibly alleged activity in connection with a “commercial transaction” or in “trade or commerce.” (See supra Section V.C.) However, the Court must separately consider the actionability of any affirmative representations or omissions alleged by YouTube, Snap, and TikTok.

In Count 7 of the 2AMC, plaintiffs allege, against all defendants, that they committed fraudulent, unfair, deceptive and/or abusive practices amounting to state consumer-protection violations including:

(a) Representing that goods or services have approval, characteristics, uses, or benefits that they do not have;

(b) advertising goods or services with the intent not to sell them as advertised;

(c) engaging in deception, fraud, false pretense, false premise, misrepresentation, or knowing concealment, suppression, or omission of any material fact with the intent that a consumer rely on the same in connection with the promotion of products and/or that creates a likelihood of confusion; and

(d) Failing to provide adequate notice to parents about collecting personal information of children under the age of 13, as required by COPPA and accompanying regulations, the violation of which “shall be treated as a violation of a rule defining an unfair or deceptive act or practice.” 15 U.S.C. § 6502(c).

(2AMC ¶ 962.) The gravamen of plaintiffs' consumer-protection theory as rooted in defendants' “false and misleading misrepresentations or omissions of material fact relating to the safety of its respective social media products.” (Id. ¶ 964.)⁹⁴

Defendants note that the 2AMC lacks any allegations of affirmative statements by TikTok or YouTube. According to defendants, the 2AMC attributes one affirmative statement to Snap: “Snap's Vice President of Global Public Policy, Jennifer Stout, stated in written testimony to a Senate Subcommittee that Snap takes ‘into account the unique sensitivities and considerations of minors when we design products' when, in fact, Snap intentionally designed its product to promote compulsive and excessive use ․” (Id. ¶ 540 (footnote omitted).) Plaintiffs provide no direct response to these points in their opposition.⁹⁵ On the other hand, as evident from the Court's prior order, the complaint is replete with allegations as to each of these defendants' deliberate design of platform features that they knew or should have known would pose unique risks of harm to children. See Social Media I, 702 F. Supp. 3d at 818–22.

The YouTube, Snap, and TikTok defendants' joinder to the motion to dismiss is GRANTED insofar as plaintiffs' consumer-protection claims are based on these defendants' affirmative statements. The standalone affirmative statement alleged against Snap is, without more, a nonactionable statement of opinion. See In re Yahoo! Inc. Customer Data Sec. Breach Litig., 2017 WL 3727318, at *26. The joinder is DENIED as to plaintiffs' omissions and failure-to-warn theory of consumer-protection violations. As discussed above with respect to Meta, because plaintiffs' consumer-protection claims against YouTube, Snap, and TikTok as limited proceeds solely under a failure-to-warn theory, the theory is not at this stage limited by Section 230.

C. Personal Injury Plaintiffs' Concealment and Misrepresentation Claims against Meta Only

Meta moves to dismiss the personal injury plaintiffs' claims of fraudulent concealment and misrepresentation (Count 8) and negligent concealment and misrepresentation (Count 9). These are not asserted against the other MDL defendants.

“While some components of a fraud claim vary from state to state, ‘the fundamental elements of fraud are substantially similar from state to state.’ ” In re Soc. Media Adolescent Addiction/Pers. Inj. Prod. Liab. Litig., 2024 WL 1786290, at *4 (N.D. Cal. Apr. 15, 2024) (quoting In re Takata Airbag Prod. Liab. Litig., 677 F. Supp. 3d 1311, 1330 (S.D. Fla. 2023)). “Generally, the state fraudulent misrepresentation claims at issue here require plaintiffs to show (1) a false representation, (2) the speaker's knowledge of its falsity, (3) an intent to defraud, (4) the hearer's reasonable or justifiable reliance on the statement, (5) and consequent and proximate injury.” Id.

“As to claims of negligent misrepresentation, again with some variation, the core elements include: (1) a false representation, (2) the speaker's intent to induce the hearer's reliance on the statement, (3) the speaker's failure to exercise reasonable care in communicating the information (i.e., breach of a duty of care), (4) the hearer's reasonable or justifiable reliance, and (5) consequent and proximate injury.” Id.

Meta reincorporates many of its earlier arguments against the States as to the personal injury plaintiffs' concealment and misrepresentation claims, and further levies arguments specific to claims of negligent misrepresentation by omission. The Court considers each set in turn.

1. Section 230, Statements of Opinion, Noerr-Pennington, Reliance, Falsity, and Materiality

As above, Meta reincorporates by reference prior arguments levied against the States' claims. The Court's reasoning with respect to the consumer protection claims applies to the plaintiffs' fraudulent and negligent concealment and misrepresentation claims are similarly incorporated herein: (i) allegations based on an omission or failure to warn of harmful platform features are not at this stage barred by Section 230 (see supra Section IV); (ii) plaintiffs plead actionable statements by Meta (see supra Section V.A.1); (iii) the Noerr-Pennington doctrine is no bar on the facts alleged (see supra Section V.A.4); and the sufficiency of allegations for fraud (iv) as to reliance and falsity (see supra Sections V.A.2, VI.A.2); and (v) materiality (see supra Section V.A.3).

The Court GRANTS in part and DENIES in part the Meta's motion to dismiss the personal injury plaintiffs' fraudulent and negligent concealment and misrepresentation claims to the same extent as assessed above with respect to the States.

2. Negligent Misrepresentation by Omission: Whether Recognized by Any At-Issue States

Next, Meta asserts that at least eighteen states do not recognize a claim for negligent misrepresentation by omission: California, Colorado, Georgia, Hawai'i, Kansas, Kentucky, Maryland, Michigan, Nebraska, Nevada, North Carolina, Ohio, Oregon, South Carolina, Texas, Virginia, Washington, and Wisconsin. (Dkt. No. 662 at 39; Dkt. No. 662-1 at 25–27.)

“When a federal court sitting in diversity is presented with a novel issue of state law, the federal court must consider how that state's highest court would likely decide the issue.” In re Packaged Seafood Prod. Antitrust Litig., 635 F. Supp. 3d 1061, 1071 (S.D. Cal. 2022). The Court considers how each state's high court might decide this issue.

a) States Prohibiting or Unlikely to Permit Negligent Misrepresentation by Omission

The Court has already preliminarily determined that Georgia, Maryland, North Carolina, and Ohio do not or are not likely to recognize claims for negligent misrepresentation by omission. In re Soc. Media Adolescent Addiction/Pers. Inj. Prod. Liab. Litig., 2024 WL 1786290, at *6 & n.10.⁹⁶ The Court further determines that negligent misrepresentation by omission is not recognized in California, Kentucky, and Washington. The caselaw is cited below.⁹⁷ The motion is granted as to those seven states.

The Court previously discussed the availability of negligent misrepresentation by omission under Colorado law, and now determines such claims are not likely viable under that state's law. Previously, the Court noted that Colorado courts disagree as to whether negligent misrepresentation by omission can lead to liability. In re Soc. Media Adolescent Addiction/Pers. Inj. Prod. Liab. Litig., 2024 WL 1786290, at *6 & n.12; compare, e.g., Craig Hosp. v. Tyson Foods, Inc., No. 17-cv-02534, 2019 WL 5095737, at *7 (D. Colo. July 22, 2019) (collecting cases) (“[T]he Court concludes that the Colorado Supreme Court would require that a negligent misrepresentation claim be grounded in affirmative statements.”), with Elevation 10175, LLC v. Douglas County Ins. Services, Inc., No. 19CV30015, 2020 WL 11271852, at *3 (Colo. Dist. Ct. Dec. 01, 2020) (“Though such caselaw is sparse, the court finds sufficient support in the opinions of higher courts to conclude that the negligent omission of information can lead to liability in Colorado.”). While these Colorado appellate courts have evidently not addressed the question further, one federal district court in Colorado confronted with this question has consistently held that “the Colorado Supreme Court would require that a negligent misrepresentation claim be grounded in affirmative statement.” Cahey v. Int'l Bus. Machines Corp., No. 20-cv-00781, 2020 WL 5203787, at *8 (D. Colo. Sept. 1, 2020); Licerio v. Lamb, No. 20-cv-00681, 2021 WL 4556092, at *23 n.42 (D. Colo. July 15, 2021), report and recommendation adopted, 2021 WL 4167341 (D. Colo. Sept. 14, 2021), aff'd, No. 21-1375, 2022 WL 4100400 (10th Cir. Sept. 8, 2022); Martin v. Chinese Child. Adoption Int'l, No. 19-cv-02305, 2020 WL 6585796, at *13–14 (D. Colo. Nov. 10, 2020). Accordingly, this Court declines to permit negligent misrepresentation claims not grounded in affirmative statement under Colorado law.

The motion is also granted as to claims of negligent misrepresentation by omission under Colorado law.

b) States with Sufficient Authority to Permit Negligent Misrepresentation by Omission

The Court preliminarily determines that sufficient authority exists to permit claims of negligent misrepresentation by omission to proceed under the laws of Michigan and Nevada.

Michigan. The Michigan Court of Appeals has “declined to extend the tort of negligent misrepresentation beyond the misrepresentation of facts that can be independently verified.” The Mable Cleary Tr. v. The Edward-Marlah Muzyl Tr., 686 N.W.2d 770, 783 (Mich. Ct. App. 2004), overruled on other grounds by Titan Ins. Co. v. Hyten, 817 N.W.2d 562, 568 n.4 (Mich. 2012). Meta cites caselaw which asserts this proposition from The Mable Cleary Tr. means that “a claim for negligent misrepresentation must be based on misrepresentations of fact, not on opinions or omissions.” Donald v. Hi-Tec Bldg. Servs., Inc., 2013 WL 5521632, at *2 (W.D. Mich. Oct. 3, 2013). Plaintiffs respond with citation to Burket, which considered this interpretation “misplaced” and noted in particular that (1) the Supreme Court of Michigan in Williams v. Polgar permitted liability for negligent misrepresentation based on an omission, and (2) the at-issue omission (an individual's access to certain bank accounts) is a fact that can be independently verified. Burket v. Hyman Lippitt, P.C., 560 F. Supp. 2d 571, 597 & n.22 (E.D. Mich. 2008) (discussing Williams v. Polgar, 215 N.W.2d 149, 154–57 (Mich. 1974)).⁹⁸ “It follows then that a claim for negligent misrepresentation under Michigan law can be based on omissions so long as there is a duty to disclose.” Id. The Court agrees with Burket's assessment and permits plaintiffs' claims of negligent misrepresentation by omission under Michigan law to proceed.

Nevada. The Supreme Court of Nevada has affirmed summary judgment on a “negligent misrepresentation claim because” the defendant “neither made an affirmative false statement nor omitted a material fact it was bound to disclose.” Noonan v. Bayview Loan Servicing, LLC, 438 P.3d 335 (Nev. 2019). While there is potential for a more nuanced construction of this statement, because Meta is unable to provide further clarity at this stage, this Court will permit the plaintiffs' claims of negligent misrepresentation by omission under Nevada law to proceed.

c) States Without Clear Authority as to Negligent Misrepresentation by Omission

As to Hawai'i, Kansas, Oregon, Nebraska, South Carolina, Texas,⁹⁹ Virginia, and Wisconsin, the parties' cited cases provide an insufficient basis to determine whether the supreme courts of those states would either endorse or reject a claim of negligent misrepresentation by omission.¹⁰⁰ See In re Packaged Seafood Prod. Antitrust Litig., 635 F. Supp. 3d at 1071.

For instance, consider Nebraska. In Gibb v. Citicorp Mortg., Inc., the Supreme Court of Nebraska adopted the tort of negligent misrepresentation as defined in Restatement (Second) of Torts § 552 (1977). Creighton Univ. v. Gen. Elec. Co., 2009 WL 756206, at *3 (D. Neb. Mar. 18, 2009) (citing Gibb v. Citicorp Mortg., Inc., 518 N.W.2d 910, 921 (Neb. 1994)). “One may be liable for negligent misrepresentation ‘if one fails to exercise the level of care required under the circumstances.’ ” Cargill v. Kroeger, 2012 WL 12884553, at *8 (D. Neb. Aug. 15, 2012) (quoting Gibb, 518 N.W.2d at 921).

In support of its position that Nebraska does not recognize negligent misrepresentation by omission, Meta cites to a Supreme Court of Nebraska case predating Gibb which evaluated the scienter applicable to a fraudulent concealment claim. Nelson v. Cheney, 401 N.W.2d 472, 476 (Neb. 1987). The Nelson court discussed a prior case “in which [the court] determined that ‘intent to deceive’ was not a necessary element of a cause of action for false representation.” Id. (citing Nielsen v. Adams, 388 N.W.2d 840 (Neb. 1986)). Instead, the court had permitted a misstatement of a material fact made “recklessly, without having proper knowledge from a reasonable inquiry,” to support a cause of action for false representation. Id. Nonetheless, the Nelson court determined that “we do not believe it wise or good public policy to adopt a theory of negligent concealment in the vendor/purchaser setting.” Id.

The Nelson court's decision is more nuanced than Meta suggests and explicitly limits its reference to “negligent concealment” to its circumstances. Absent a more tailored showing of how the Supreme Court of Nebraska might approach this question, the Court declines to dismiss personal injury plaintiffs' negligent misrepresentation by omission claims under Nebraska law at this stage.

Because the burden on a motion to dismiss falls to the moving party, and because plaintiffs' claims of negligent misrepresentation by omission are not dispositive of their actions, plaintiffs' claims of negligent misrepresentation by omission under the laws of Hawai'i, Kansas, Nebraska, South Carolina, Virginia, and Wisconsin survive Meta's motion to dismiss. The parties will have an opportunity to provide a more nuanced view of how those states' supreme courts would approach this question as to any applicable bellwether plaintiffs on summary judgment.

Meta's motion to dismiss personal injury plaintiffs' claims of negligent misrepresentation by omission is GRANTED as to plaintiffs bringing claims under the laws of California, Colorado, Georgia, Kentucky, Maryland, North Carolina, Ohio, and Washington, and DENIED as to the laws of other states.

VII. CONCLUSION

To summarize, Meta's motion to dismiss is largely DENIED.¹⁰¹ The Court outlines here the scope limitations set forth in this order.

• States' COPPA Claims

The motion is DENIED. Two theories can be asserted for this claim. With respect to the first, Meta concedes that the States' have alleged actual knowledge of under-13 users. With respect to the second, the Court HOLDS that third-party content hosted by a platform can be considered when determining whether the platform, or a portion thereof, is “directed to children” under COPPA.

• Section 230

The motion is GRANTED IN PART and DENIED IN PART utilizing the same approach the Court employed in Social Media I, 702 F. Supp. 3d at 824–35. Meta did not seek to move to dismiss the States' COPPA claim nor the States' consumer-protection claims predicated on Meta's allegedly deceptive affirmative representations as barred by Section 230, and thus, the claim proceeds.

The motion to dismiss is GRANTED to the extent the States' theories of unfair and unconscionable acts and practices target the following platform features: infinite scroll and autoplay; ephemeral content; how Meta designed and deployed audiovisual and vibration notifications and alerts; the quantification and display of likes; how Meta algorithmically served content to young users.

The motion to dismiss is DENIED as to the States' theories of unfair and unconscionable acts and practices concerning appearance altering filters; platform features that hindered time restrictions; and Instagram's “multiple accounts” function.

The motion to dismiss is DENIED insofar as the states' claims and the personal injury plaintiffs' consumer-protection, misrepresentation, and concealment theories seek to impose liability for a failure to warn of known risks of addiction attendant to platform features or as to platform construction in general.

• Consumer Protection Claims

○ Theory Based on Deceptive Acts and Practices

The motion to dismiss is DENIED. The States have alleged at least some material representations that, drawing all reasonable inferences in the States' favor, could be considered deceptive by a factfinder. The Court declines to parse and strike individual statements at the motion-to-dismiss stage. Currently, nothing in the record suggests that Meta's statements to Congress implicate its First Amendment right to petition and merit protection under the Noerr-Pennington doctrine.

○ Theory Based on Unfair and/or Unconscionable Acts and Practices

The motion to dismiss is DENIED. Plaintiffs' have presented sufficient allegations under various tests, including Section 5 of the FTC Act, the Sperry factors, and the various state-specific considerations Meta has raised.

○ Scope of Consumer Protection Statutes

The motion to dismiss on the grounds that certain states' statutory language provide for a narrower scope than is alleged is DENIED. The States and personal injury plaintiffs have adequately alleged conduct occurring in connection with a “consumer transaction,” or in “trade or commerce” as specific states' laws require. Additionally, Meta has not sustained its burden as to whether the personal injury plaintiffs asserting claims arising under Alabama, Texas, and Vermont are “consumers” under those states' laws as to permit a private right of action for unfair and deceptive acts and practices.

• Restitution

The motion to dismiss is DENIED. Given state-specific nuances as to the scope of available restitution under consumer-protection claims brought by state attorneys general, Meta's approach cuts too broadly at this preliminary stage.

• Personal Injury Plaintiffs' Claims as to Meta

○ Consumer-Protection, Misrepresentation, and Concealment Claims

The motion to dismiss is GRANTED IN PART and DENIED IN PART largely to the same extent assessed as to the States' claims on the issues of Section 230, statements of opinion, and the Noerr-Pennington doctrine. With respect to arguments of reliance:

The motion to dismiss is GRANTED under the laws of Arizona, California, Georgia, Indiana, Iowa, Nevada, North Carolina, Pennsylvania, South Dakota, Texas, and Virginia for claims predicated on affirmative statements for failure to plead reliance;

The motion to dismiss is DENIED under the laws of Colorado, Maine, Massachusetts, Minnesota, New Jersey, North Dakota, Ohio, Oregon, Rhode Island, and South Carolina for claims predicated on affirmative statements; and

The motion to dismiss is DENIED insofar as plaintiffs seek to impose liability on defendants' omissions or failures to disclose under the law of any at-issue state.

○ Negligent Misrepresentation by Omission

The motion to dismiss is GRANTED as to plaintiffs bringing claims under the laws of California, Colorado, Georgia, Kentucky, Maryland, North Carolina, Ohio, and Washington and DENIED as to the laws of other states.

• Personal Injury Plaintiffs' Consumer Protection Claims as to YouTube, Snap, and TikTok

The personal injury plaintiffs' consumer-protection claims against YouTube, Snap, and TikTok are GRANTED to the extent based on affirmative statements and DENIED to the extent based on an omissions or failure-to-warn theory.

In short, the States' consumer-protection claims against Meta under the unfair and unconscionable acts and practices prong that target the design, development, and deployment of three platform features—appearance altering filters, features hindering time restrictions, and the “multiple accounts” function—survive as to all States so asserting. The States' claims against Meta under the deceptive acts and practices prong survive as to all affirmative statements, and as to a failure to warn of known risks of addiction attendant to platform features or as to platform construction in general.

The personal injury plaintiffs' consumer-protection, concealment, and misrepresentation claims against Meta are limited to the same extent as the States' deceptive acts and practices claims, and additionally (i) to the extent predicated on affirmative statements, are dismissed where arising under the laws of Arizona, California, Georgia, Indiana, Iowa, Nevada, North Carolina, Pennsylvania, South Dakota, Texas, and Virginia, (ii) survive to the extent such claims seek to impose liability under a “pure” omissions or failure-to-warn theory under the law of any state, (iii) except claims of negligent misrepresentation by omission are dismissed under the laws of California, Colorado, Georgia, Kentucky, Maryland, North Carolina, Ohio, and Washington.

The personal injury plaintiffs' consumer-protection claims against YouTube, Snap, and TikTok survive only to the extent based on an omissions or failure-to-warn theory.

This terminates Dkt. No. 517 in Case No. 22-md-03047; Dkt. No. 83 in Case No. 23-cv-05448; and Dkt. No. 15 in Case No. 23-cv-05885.

IT IS SO ORDERED.

ATTACHMENT A

TABLE OF CONTENTS

I. Background A. Procedural Background B. Relevant Facts Alleged 1. Meta's Alleged Design and Deployment of Harmful Platform Features Targeting Young Users 2. Meta's Alleged Awareness and Concealment of Harmful Platform Features From the Public 3. COPPA Allegations C. Theories of Liability 1. Deceptive Acts and Practices 2. Unfair and/or Unconscionable Acts and Practices II. Legal Framework III. The Children's Online Privacy Protection Act IV. Section 230 A. Overview B. Legal Framework C. Unfair and Unconscionable Acts: Platform Development and Design 1. Features Previously Barred by Section 230 2. Features Previously Not Barred by Section 230 D. Deceptive Acts and Practices: Failure-to-Warn Theory V. Consumer Protection Claims against Meta A. Deceptive Acts and Practices 1. Puffery and Opinion 2. Sufficiency of Pleading Falsity under Rule 9(b) 3. Materiality of Alleged Deceptions 4. Noerr-Pennington B. Unfair and/or Unconscionable Business Acts or Practices 1. Section 5 of the FTC Act 2. The Sperry Factors a) Whether Sperry Applies b) Applying the Sperry Factors 3. State-Specific Standards a) California b) Colorado c) Indiana d) Kansas e) New Jersey f) Oregon C. Statutory Scope of Consumer Protection Claims 1. “Consumer Transaction” or “Consumer Goods or Services” 2. “Trade or Commerce” 3. “Consumer” D. Restitution VI. Personal Injury Plaintiffs' Claims A. Personal Injury Plaintiffs' Consumer-Protection Claims Against Meta Only 1. Prior Rulings 2. Actual Reliance a) Disputed State Standards b) Evaluating Actual Reliance 3. Ascertainable Loss B. Personal Injury Plaintiffs' Consumer-Protection Claims Against Snap, YouTube, and TikTok C. Personal Injury Plaintiffs' Concealment and Misrepresentation Claims against Meta Only 1. Section 230, Statements of Opinion, Noerr-Pennington, Reliance, Falsity, and Materiality 2. Negligent Misrepresentation by Omission: Whether Recognized by Any At-Issue States a) States Prohibiting or Unlikely to Permit Negligent Misrepresentation by Omission b) States with Sufficient Authority to Permit Negligent Misrepresentation by Omission c) States Without Clear Authority as to Negligent Misrepresentation by Omission VII. Conclusion 34 4 8 12 13 13 14 15 16 2 21 22 24 25 29 31 37 37 37 40 43 44 45 46 50 50 51 52 52 55 57 58 59 60 62 62 67 71 73 77 77 77 78 78 82 84 87 89 89 90 90 92 93 96

FOOTNOTES

1.   The entities include Meta Platforms, Inc. (Meta Platforms), Instagram, LLC (Instagram), Meta Payments, Inc. (Meta Payments), and Meta Platforms Technologies, LLC (Meta Technologies), which the Court refers to collectively as “Meta.” Multistate Compl. ¶ 22.

2.   For the reader's ease, and given the length of the order, a table of contents is included as Attachment A.

3.   References to the “States” include all 34 state attorneys general who assert claims in the multistate action and the Florida AG's action unless otherwise specified.On February 9, 2024, the Judicial Panel on Multidistrict Litigation transferred the state of Montana's action against Meta defendants into this MDL. Dkt. No. 605; see also Montana v. Meta Platforms, Inc., No. 23-cv-00145 (D. Mont.); Montana v. Meta Platforms, Inc., No. 24-cv-00805 (N.D. Cal.). Montana is thus not included in any references to the “States” herein. The impact of this order on Montana's complaint shall be addressed at the next conference.

4.   The personal injury plaintiffs initially filed a first amended complaint on April 14, 2023, see Dkt. No. 234-1, and did not (because they could not) incorporate any allegations of the Multistate Complaint, which was filed on October 24, 2023. The 2AMC was filed on November 15, 2023, and adopts by reference many of the new allegations presented in the Multistate Complaint. See 2AMC, ¶ 391A. Because many of the allegations uniquely presented in the States' complaint have been incorporated by reference into the 2AMC, much of the new relevant factual material described below applies to the personal injury plaintiffs' claims.

5.   For the Court's discussion of these features as alleged by the personal injury plaintiffs, see Social Media I, 702 F. Supp. 3d at 819–22. Given the overlap in material discussed in the personal injury plaintiffs' complaint, any additional details regarding the platform features from that discussion not repeated in this Order are considered incorporated herein.

6.   The Multistate Complaint also discusses video features likes “IGTV,” “Instagram Video,” and “Reels.” (Multistate Compl. ¶¶ 107–16.)

7.   Under 15 U.S.C. § 6504, the States request as relief to: (i) enjoin conduct violating COPPA and the COPPA Rule, (ii) enforce compliance with the COPPA Rule, (iii) obtain damages, restitution, and other compensation, and (iv) any other relief as the Court may consider appropriate. (Multistate Compl. ¶ 858; see also Florida Am. Compl. at 34.)

8.   The States also allege that “Meta has made other false and deceptive representations, including as set forth in paragraphs 1 through 835.” (Multistate Compl. ¶ 846(g).)

9.   As discussed infra, the FTC has promulgated regulations in 16 C.F.R. §§ 312.1–312.13 that in part interpret COPPA's provisions and in part implement certain procedures required by section 6502(b) of COPPA.

10.   Meta's reply is unavailing. First, Meta subtly misquotes the statute to support its construction that the states allege two separate claims. Dkt. No. 662 at 1. Second, Meta's sole case cited in opposition, Cesnik v. Edgewood Baptist Church, 88 F.3d 902, 905 (11th Cir. 1996), concerned a complaint which pled “at least nine discrete theories of recovery” in one count, e.g., breach of fiduciary duty and misrepresentation and intentional infliction of emotional distress, among others.

11.   As noted in the text of the rule, each of the requirements in subsections (a), (b), (c), (d), and (e) are separately elaborated in sections 312.4(b), 312.5, 312.6, 312.7, and 312.8 of the Rule, respectively.

12.   Meta requests incorporation by reference and judicial notice of the third-party Instagram accounts linked in the Multistate Complaint and video advertisements excerpted in the Multistate Complaint. Dkt. No. 517-4. Because the Court does not here engage with the “directed to children” analysis that those materials implicate, Meta's request is DENIED as unnecessary.

13.   FAQ D.2 says that (i) a third-party provider posting child-directed content for a commercial purpose on a general audience platform, to the extent it collects personal data, will need to comply with COPPA, and (ii) if the platform has “actual knowledge that [the third-party's] content is directed to children and is collecting personal information, it will also need to comply with COPPA.”

14.   A blanket exclusion of considering third-party content also seems inconsistent with recent guidance from the FTC on its “mixed audience” exception. See Notice of Proposed Rulemaking, 89 Fed Reg. 2034, 2039 (Jan. 11, 2024) (“For example, where third-party content on a platform is child-directed under the Rule's multi-factor test but the platform does not target children as its primary audience, the operator can request age information and provide COPPA protections only to those users who are under 13.”).

15.   Notably, the court in Get Oil Out! rendered a decision at summary judgment, not on a motion to dismiss.

16.   In enacting Section 230, Congress “sought to encourage the development and use of technologies that would allow users to filter and control the content seen by themselves or their children.” Est. of Bride by & through Bride v. Yolo Techs., Inc., 112 F.4th 1168, 1175 (9th Cir. 2024); see also 47 U.S.C. § 230(b)(4) (“It is the policy of the United States,” among other goals, “to remove disincentives for the development and utilization of blocking and filtering technologies that empower parents to restrict their children's access to objectionable or inappropriate online material.”).

17.   In its reply, Meta also argues that a child-directed “portion” of a website must be “created by the service” itself (Dkt. No. 662 at 5–6 n.6); that is, third-party account pages cannot constitute “portions” of the website. This restriction is again nowhere present in the statute or Rule, and Meta provides inapposite citations to the congressional record and certain regulations in support. See, e.g., 144 Cong. Rec. S12741-04 (if a “site has a special area for children, then that portion of the site will be considered to be directed to children”); 64 Fed. Reg. 59893 (“[I]f a general audience site has a distinct children's ‘portion’ or ‘area,’ then the operator would be required to provide the protections of the Rule for visitors to that portion of the site.”). Neither citation is inconsistent with this Court's holding.

18.   The States confirmed at argument that they “seek[ ] to impose liability on Meta's failure to speak about the overall effect of its unfair business practices.” Dkt. No. 785, Tr. of April 19, 2024 Case Management Conference at 163:4–10.

19.   See also Google, Inc. v. Hood, 822 F.3d 212, 227 n.12 (5th Cir. 2016) (noting the court “do[es] not suggest that section 230 of the CDA would not apply if [Attorney General] Hood were to eventually bring an enforcement action or cannot be applied at the motion-to-dismiss stage.”).

20.   47 U.S.C. § 230(e)(3) provides, in part, that “[n]o cause of action may be brought and no liability may be imposed under any State or local law that is inconsistent with this section.”

21.   In its prior order, the Court held that allegations under a products liability theory targeting certain features of the defendants' platforms were insulated by Section 230, while other features were not so insulated. See Social Media I, 702 F. Supp. 3d at 829–35. While consumer-protection claims of unconscionability and unfairness present a distinct legal framework as compared to products liability, the Court finds that its prior analysis of Meta's platform features under a products liability theory leads to the same result under the States' unconscionability and unfairness theories for the purposes of Section 230.

22.   While not at issue in Meta's motion to dismiss, the Court has also held that “the timing and clustering of notifications of defendants' content to increase addictive use is entitled to First Amendment protection.” Social Media I, 702 F. Supp. 3d at 837.

23.   To the extent the States argue that Section 230 should not apply because Meta wrongfully knew its design choice would lead to “negative social comparison,” at least one court has explained that Section 230 does not contain an exception for wrongful intent. See Levitt v. Yelp! Inc., No. C-10-1321, 2011 WL 5079526, at *1 (N.D. Cal. Oct. 26, 2011) (“Indeed, courts have found the [Subsection 230(c)(1)] immunity applies to conduct that arguably constitute[s] bad faith.”), aff'd, 765 F.3d 1123 (9th Cir. 2014). While perhaps unseemly, this Court agrees from a legal perspective.

24.   “These algorithms do not promote any specific message by Meta. Rather, the algorithms function on a user-by-user basis, detecting the material with which each individual is likely to engage and then increasingly displaying similar material to maximize the time spent (and user data collected) on the Platforms.” Multistate Compl. ¶ 65.

25.   The States also argue that the Ninth Circuit has not expanded the meaning of publication to include “promoting, marketing, or boosting content,” as allegedly done by Meta's algorithms. These activities involve the republication of content, and thus directly implicate decisions of “whether to publish or to withdraw from publication third-party content.” Barnes, 570 F.3d at 1102.

26.   Of the three product features discussed in this section, all were discussed in the Court's prior order except for the last, Instagram's “multiple accounts” function, which the Court adds here given its ultimate finding on similar grounds.

27.   The States argue Meta failed to move to dismiss this feature. Meta argues that the complaint did not provide sufficient notice. While this feature was not included in the complaint's summary list, see Multistate Complaint ¶ 847, the States have included a substantial section dedicated to appearance-altering filters, see id. ¶¶ 333–68. Regardless, because the Court has ruled on similar allegations previously and Meta has not provided a new basis for dismissal, there is no prejudice to ruling on the issue now.

28.   The parties again dispute the articulation of a claim and a ground for a motion. The Court again rules for efficiency purposes and lack of prejudice.

29.   The States argue again that Meta has not moved to dismiss and so concedes that Section 230 does not bar allegations related to Instagram's “multiple accounts” function. See Multistate Compl. ¶¶ 370–72. In the interest of efficiency, the Court considers the argument now, finding sufficient notice of allegations despite Meta's objection (id. ¶¶ 369–72) and sufficient briefing on the legal issues despite the States' objection.

30.   Meta confirmed at argument. Dkt. No. 785, Tr. of April 19, 2024 Case Management Conference at 68:10–16; see also Dkt. No. 517 at 19.

31.   See also In re Facebook, Inc., 625 S.W.3d 80, 93 (Tex. 2021) (dismissed because “duty to warn” users “against recruitment into sex trafficking by other users” treated Facebook “ ‘as the publisher or speaker’ of third-party communication”).But see A.M. v. Omegle.com, LLC, 614 F. Supp. 3d 814, 817 (D. Or. 2022) (claim of “fail[ure] to warn child users of adult predators on the website” was not barred by Section 230 because “Omegle would not have to alter the content posted by its users,” only “its design and warnings”).

32.   See also Bride, 112 F.4th at 1181 (“[T]he defendant in Internet Brands failed to warn of a known conspiracy operating independent of the site's publishing function.”); Herrick, 306 F. Supp. 3d at 592 (“Internet Brands is best read as holding that the CDA does not immunize an [interactive computer service] from a failure to warn claim when the alleged duty to warn arises from something other than user-generated content.”).

33.   The court in Herrick further clarified that it was “persuaded that requiring Grindr to post a warning at the outset or along with each profile is no different than requiring Grindr to edit the third-party content itself.” Herrick, 306 F. Supp. 3d at 591–92. That is, whether a platform displayed content warnings on individual posts or profiles versus one global content warning upon entering the platform made no difference for the purposes of Section 230.

34.   Copies of these opinions which lack a database citation were filed on this MDL's docket at Dkt. Nos. 952 (Arkansas), 1045-2 (Vermont), 1045-3 (Utah), 1045-4 (New Mexico), and 1140-1 (D.C.).

35.   In Anderson v. TikTok, Inc., the Third Circuit held that TikTok's recommendation algorithm “is TikTok's own expressive activity,” thus falling outside Section 230 and instead within the First Amendment's coverage. No. 22-3061, 2024 WL 3948248 (3d Cir. Aug. 27, 2024). The court reached this issue based on its read of the Supreme Court's opinion in Moody, in which the “Court held that a platform's algorithm that reflects ‘editorial judgments’ about ‘compiling the third-party speech it wants in the way it wants’ is the platform's own ‘expressive product’ and is therefore protected by the First Amendment.” Id. (quoting Moody v. NetChoice, LLC, 144 S. Ct. 2383, 2394 (2024)).

36.   Meta's motion focuses on the actionability of its affirmative statements. The States' deceptive acts and practices claim, as discussed above, also relies on any omissions and failures to warn. As such, any omissions and failures to warn survive Meta's motion to dismiss on the merits, consistent with this Court's findings as to affirmative statements.

37.   Meta included as an attachment to its motion to dismiss a 22-page chart listing misrepresentations alleged in the Multistate Complaint, marking which of the four grounds of dismissal applies to each statement. Dkt. No. 517-1.

38.   The standard for common law deception is much less exacting than claims arising under the Private Securities Litigation Reform Act (“PSLRA”). Thus, those cases cited by Meta rooted in the securities context do not persuade. See, e.g., In re Lyft Inc. Sec. Litig., 484 F. Supp. 3d 758, 770 (N.D. Cal. 2020).Meta also relies on Newcal Indus. Inc. v. Ikon Office Sol., in which the Ninth Circuit, assessing misrepresentations under the Lanham Act, wrote that “a statement that is quantifiable, that makes a claim as to the ‘specific or absolute characteristics of a product,’ may be an actionable statement of fact while a general, subjective claim about a product is non-actionable puffery.” 513 F.3d 1038, 1053 (9th Cir. 2008). While this general proposition may be applicable, the Court does not agree that Newcal's statement-by-statement assessment of the five representations at issue in that case provides the appropriate analytical framework for the instant consumer-protection and, as discussed later, common-law fraud causes of action.

39.   Meta explained on other occasions that Project Daisy was not implemented because it was “beneficial for some, and annoying to others,” and that it “ultimately didn't have as strong an impact as we'd hoped,” among other statements. See Multistate Compl. ¶¶ 284–87. While these statements are facially more nuanced, the Court need only determine whether the States have plausibly alleged an actionable misrepresentation. Further assessment skirts too closely to deciding disputes of fact that would be inappropriate on a 12(b)(6).

40.   Certain states dispute whether Rule 9 applies to deceptive acts and practices claims, arguing that the Ninth Circuit has questioned whether Rule 9(b) requirements apply to violations of consumer protection statutes. Compare Moore v. Mars Petcare US, Inc., 966 F.3d 1007, 1019 (9th Cir. 2020) (“Rule 9(b) requirements may not even be necessary, given that a defendant can violate the [Unfair Competition Law], [False Advertising Law], and [Consumer Legal Remedies Act] by acting with mere negligence.”), with Avakian v. Wells Fargo Bank, N.A., 827 F. App'x 765, 766 (9th Cir. 2020) (“[C]laims for unfair business practices, intentional misrepresentation, fraud, and negligent misrepresentation are all fraud-based claims, and must meet the heightened pleading requirements of Rule 9(b).”). Because the Court determines that the States' allegations satisfy Rule 9, it need not reach the issue.

41.   Meta emphasizes that, under Rule 9(b), the plaintiff must show the “mismatch between the representations at issue and the evidence that allegedly debunks them.” Amado v. Procter & Gamble Co., No. 22-cv-05427, 2023 WL 3898984, at *9 (N.D. Cal. June 8, 2023). The Court disagrees that the “mismatch” need be as stringent as Meta urges. First, while a court may determine that a statement is not deceptive as a matter of law, the determination is generally left to the factfinder. Second, the standard for evaluating “[w]hether a business practice is deceptive or misleading” under, for example, California's Unfair Competition Law (“UCL”), is “governed by the ‘reasonable consumer’ test.” Moore v. Mars Petcare US, Inc., 966 F.3d 1007, 1017 (9th Cir. 2020) (quoting Williams v. Gerber Prod. Co., 552 F.3d 934, 938 (9th Cir. 2008)). The case from which this “mismatch” language appears to originate, Eckler v. Wal-Mart Stores, Inc., was itself assessing allegations under the “reasonable consumer” test and noted in a descriptive manner that inadequately pled false advertising claims lack “this kind of mismatch” between the representations and evidence allegedly debunking the false representations. No. 12-cv-727, 2012 WL 5382218, at *6 (S.D. Cal. Nov. 1, 2012). Meta's strict application of “mismatch” would serve to heighten the “reasonable consumer” standard and advance that fact-focused inquiry into the motion-to-dismiss stage.

42.   “A misrepresentation is judged to be ‘material’ if ‘a reasonable [person] would attach importance to its existence or nonexistence in determining his choice of action in the transaction in question.’ ” Id. (quoting Engalla v. Permanente Med. Grp., Inc., 938 P.2d 903, 919 (Cal. 1997))); State v. Weinschenk, 868 A.2d 200, 206 (Me. 2005) (representation considered material if it “involves information that is important to consumers and, hence, likely to affect their choice of, or conduct regarding, a product” (quoting In re Cliffdale Assocs., Inc., 103 F.T.C. 110, 165 (1984)); Hussain v. Citizens Fin. Grp., Inc., 2019 WL 490091, at *5 (N.J. Super. Ct. App. Div. Feb. 8, 2019) (“A fact is material when the seller knows or should know it is important to the particular buyer's decision or when such fact would be important to the decision of a reasonable buyer.”).

43.   Meta also argues that statements to Congress are not made within “trade or commerce” or in connection with a “sale,” “advertisement,” or “consumer transaction” as required by certain states' consumer protection laws. Meta argues that these statutory limitations should wholesale bar the States' consumer protection claims, which the Court considers infra Section V.C. Meta's arguments as to whether these statutory limitations protect its alleged misrepresentations to Congress are misplaced, as evident from the analysis in that section.

44.   FTC v. Sperry & Hutchinson Co., 405 U.S. 233 (1972).

45.   Meta cites Phillips for the proposition that “[t]his criterion is not satisfied, however, ‘if a consumer can avoid the defendant's practice by seeking an alternative elsewhere.’ ” Phillips v. Double Down Interactive LLC, 173 F. Supp. 3d 731, 743 (N.D. Ill. 2016) (quoting Batson v. Live Nation Ent., Inc., 746 F.3d 827, 833 (7th Cir. 2014)). Technically, Phillips applied the Sperry factors—not the “not reasonably avoidable” language of the FTC Act. The “criterion” that Phillips referred to is the immoral, unethical, oppressive, and unscrupulous component of the Sperry factors. Id. While analytically distinct, cases like Phillips wrestle with essentially the same problem of whether a business act or practice exploiting the psychological triggers of addiction is not reasonably avoidable.

46.   Also, as the States noted with respect to their COPPA claim, the States allege that by one 2021 estimate, of children ages 9 to 12 in the United States, 45% used Facebook and 40% used Instagram daily. (Multistate Compl. ¶ 696.) Such a widespread means for social communication among minors may not be as readily, or reasonably, avoidable.

47.   Meta does not advance feature-specific arguments with respect to allegations of appearance-altering filters.

48.   More broadly, the Court is not convinced Meta's Section 5 challenge to COPPA is appropriately framed. Unlike Meta's opposition to the States' other, feature-specific acts and practices claims, Meta's opposition to COPPA under Section 5 resembles a challenge to an FTC rulemaking. One off-hand paragraph discussing one element of Section 5 as to COPPA is insufficient to mount a challenge to a rule in place for over two decades.

49.   It is telling that Meta in its reply only gestures in conclusory fashion at the “meaningful societal benefits of Facebook and Instagram” as the countervailing benefits that outweigh these alleged harms. Dkt. No. 662 at 25–25. At the very least, this response is not appropriately tailored. The States' allegations target specific design choices that foster compulsive use, and those choices cannot be explained away by general reference to any beneficial functions Meta's platforms may serve to promote social connectivity and engagement.

50.   The Court does not endeavor to assess what standard is applicable for each of the other eighteen states Meta has identified as purportedly tracking the FTC Act. It is enough, pursuant to the above analysis, to know that both the FTC Act and the Sperry test are at issue, and that the States' claims survive under either standard at this stage.

51.   Meta also separately moves to dismiss the unfair or unconscionable acts or practices claims of Minnesota and Missouri. These states follow the Sperry factors, and so those claims survive as explained supra. See Minn. Stat. Ann. § 325F.69, subd. 8; Ward v. W. Cnty. Motor Co., 403 S.W.3d 82, 84 (Mo. 2013).

52.   Specifically, “courts held that ‘unfair’ conduct occurs when that practice ‘offends an established public policy or when the practice is immoral, unethical, oppressive, unscrupulous or substantially injurious to consumers.’ ” Davis, 691 F.3d at 1169 (citation omitted).

53.   Of note, “[t]he power to create and define an exception to the UCL is committed to the Legislature.” Aron v. U-Haul Co. of California, 49 Cal. Rptr. 3d 555, 560 (Cal. Ct. App. 2006) (citing Cel-Tech, 973 P.2d at 541 (the “safe harbor” found in legislation must clearly bar the action or permit the challenged conduct); Krumme v. Mercury Insurance Co., 123 Cal. App. 4th at 940 n.5 (“a safe harbor statute must explicitly prohibit liability for the defendant's acts or omissions.”)).

54.   The Ninth Circuit affirmed the district court's finding that NetChoice was likely to succeed in showing that one of CAADCA's reporting requirements facially violated the first Amendment, but vacated the district court's findings as to CAADCA's other mechanisms for further review. See Bonta, 2024 WL 3838423, at *7–15.

55.   Cel-Tech has traditionally applied only to claims of anticompetitive conduct. See Davis, 691 F.3d at 1170.

56.   The language of what is currently subsection (1)(rrr) appears to have been enacted on May 23, 2019. See 2019 Colo. Legis. Serv. Ch. 268 (H.B. 19-1289) (West). On May 26, 2022, the legislature moved the language to subsection (1)(rrr). 2022 Colo. Legis. Serv. Ch. 255 (H.B. 22-1287) (West).

57.   Meta does not challenge or otherwise respond to the States' contentions regarding what standard—Section 5, Sperry, or otherwise—the Colorado Supreme Court would be likely to endorse for the States' unfairness and unconscionability claim. As such, the Court need not consider the issue at this stage. The CCPA claim survives because the Court has rejected Meta's only basis for dismissal.

58.   The Court recognizes that subsection 6-1-105(1)(rrr) itemizes some unfair or unconscionable acts or practices as a form of “deceptive trade practice.” Colo. Rev. Stat. Ann. § 6-1-105(1)(rrr). The Court preliminarily considers this as a unique aspect of the CCPA's statutory organization rather than a legislative intent to proscribe claims of unfair or unconscionable acts or practices. While the States have not provided a case affirmatively endorsing the use of the CCPA for claims of unfairness or unconscionability, Meta provides no affirmative legislative or judicial determination that claims of unfair or unconscionable acts or practices are impermissible. To the contrary, Colorado caselaw is replete with references to “unfair or deceptive” practices. See, e.g., Hall, 969 P.2d at 235 (emphasis supplied). Moreover, subsection 6-1-105(3) indicates that there are “unfair trade practices actionable at common law” beyond those encompassed by subsection 6-1-105(rrr).

59.   “A deceptive act is actionable only if it is ‘incurable’ or ‘uncured.’ ” Gasbi, LLC v. Sanders, 120 N.E.3d 614, 618 (Ind. Ct. App. 2019) (quoting Ind. Code Ann. § 24-5-0.5-4(a)), vacated, 130 N.E.3d 1137 (Ind. 2019), reinstated, 130 N.E.3d 94 (Ind. 2019). “An incurable deceptive act is a deceptive act ‘done by a supplier as part of a scheme, artifice, or device with intent to defraud or mislead.’ ” Id. (quoting Ind. Code Ann. § 24-5-0.5-2(8)).

60.   Meta focuses on Crum's statement that “[t]he Court will follow the lead of the state courts by reading ‘unfair, abusive, or deceptive act’ broadly to include any act that is misleading to consumers or otherwise prohibited at law.” Crum, 2021 WL 3514153, at *2. Meta appears to urge in its motion that this language indicates the DCSA covers only (i) misleading acts or (ii) acts otherwise prohibited at law. This is a strained reading of a sentence that itself disclaims it will read the at-issue statutory terms “broadly.” Id. Meta's pivot on reply to disclaim that Indiana does not recognize “standalone unfairness claims untethered from misrepresentations or unlawful acts” admits as much. See Dkt. No. 662 at 47 (emphasis added). Meta's “tethering” of an unfair act to misrepresentations appears unfounded in caselaw other than Crum's descriptive statement and, at any rate, Meta does not attempt to explain why Indiana's DCSA claims—which target both unfair acts and deceptive representations—are not so “tethered” to Meta's alleged misrepresentations.

61.   Among the examples listed in statute, the KCPA lists as unconscionable conduct where “[t]he supplier took advantage of the inability of the consumer reasonably to protect the consumer's interests because of the consumer's physical infirmity, ignorance, illiteracy, inability to understand the language of an agreement or similar factor;” or “the supplier made a misleading statement of opinion on which the consumer was likely to rely to the consumer's detriment.” Kan. Stat. Ann. §§ 50-627(b)(1), (b)(6). Kansas contends that even if the KCPA's list of unconscionable acts was exhaustive (it is not), its allegations would qualify under either subsection (b)(1) and/or (b)(6).

62.   The States argue this misreads Kugler which acknowledged that under the CFA unconscionability is “an amorphous concept obviously designed to establish a broad business ethic,” and should be “interpreted liberally so as to effectuate public purpose.” Kugler, 279 A.2d at 651–52.

63.   Meta also argues that this claim fails because the claim has nothing to do with “selling” or “disposing” of Instagram or Facebook. This argument is considered infra Section V.C regarding statutory limitation on various States' consumer-protection statutes.

64.   In opposition to this precedent, Meta overstates the holding of In re Hard Disk Drive Suspension Assemblies Antitrust Litig., No. 19-MD-02918-MMC, 2021 WL 4306018 (N.D. Cal. Sept. 22, 2021). That court simply explained that “[a]lthough the list of ‘unconscionable tactics’ is not exclusive, there is nothing to suggest the OUTPA is intended to cover price-fixing conspiracies, statements denying price fixing, or omissions about fixed prices, conduct that is not, in any manner, similar to conduct specified in the statute.” Id.

65.   One court has written that the examples in section 646.605(9) indicate “that unconscionable trade tactics require an allegedly offending business to trick customers in some way such that the customers ‘enter into a transaction’ with the offending business.” Barber v. Unitus Cmty. Credit Union, 2023 WL 34697, at *2 (D. Or. Jan. 3, 2023) (dismissing unconscionable tactic claim because “there are no allegations that Defendants entered into a transaction with Plaintiff”). Meta reads Barber's statement out of context. Barber appears to have focused more on the fact that its pro se plaintiff failed to allege any cognizable engagement with a defendant, given his unconscionable tactics claim rested on the defendant-bank's denial of a request to open a checking account. See Amended Complaint at 1–2, 4, Barber v. Unitus Cmty. Credit Union, No. 22-cv-00134 (D. Or. June 27, 2022), ECF No. 11. Barber does not undermine the Supreme Court of Oregon's holding in Gordon that a transaction or agreement between the parties is not required to allege “unconscionable tactics.”

66.   The States bring unfair and/or deceptive acts and practices claims under Indiana, Kansas, Ohio, and Virginia law, and personal injury plaintiffs additionally bring unfair and/or deceptive acts and practices claims under Utah and Maryland law.

67.   See also Summa Health Sys. v. Viningre, 749 N.E.2d 344, 356 (Ohio Ct. App. 2000) (holding that provision of “hospital services for [plaintiff's] hysterectomy without charge due to the hospital's negligent failure to diagnose” was a consumer transaction under the Ohio Consumer Sales Protection Act (emphasis supplied)).

68.   See also State v. TikTok Inc., No. 12CV-23-65, at *8–9 (Ark. Cir. Ct. May 15, 2024) (“Although Defendants do not charge a monetary fee for the use of the TikTok app, users provide valuable consideration to TikTok—their data. Consumers agree, by bilateral contract, to provide their valuable personal data to Defendants in return for their use of the TikTok app. This contractual exchange is central to TikTok's business model. The State has adequately pleaded that the TikTok app is a ‘service’ under the” Arkansas Deceptive Trade Practices Act.); State v. Google LLC, 2022 WL 223907, at *13 (Ariz. Super. Ct. Jan. 21, 2022) (“Providing location data, in exchange for use of apps or other service[s], can certainly be considered valuable consideration under the Act.”).

69.   Before the trial court, the state argued that “TikTok users provide their personal data as consideration for access to the platform,” and that exchange qualifies as a “consumer transaction.” Indiana v. TikTok, Inc., 2023 WL 8481303, at *6. It appears the trial court did not benefit from Supreme Court of Indiana authority explaining that “Indiana has long held that consideration in the form of money is not essential to a binding contract,” and “consideration—no matter what its form—consists of a bargained-for exchange.” See Ind. Dep't of State Revenue v. Belterra Resort Ind., LLC, 935 N.E.2d 174, 179 (Ind. 2010) (citations omitted).

70.   Plaintiffs have filed a copy of the order in Utah Div. of Consumer Protection v. Meta Platforms, Inc. on the MDL docket at Dkt. No. 1045-3.

71.   Meta's caselaw is unavailing. For instance, in Matchett v. BSI Fin. Servs., the court held that a “mortgage loan, even though it involves land, is not a ‘sale, lease, ․ or other ․ transfer or disposition of goods, services, or other property’ ” and so did not constitute a “commercial transaction.” 2021 WL 3473062, at *3 (D. Utah Aug. 6, 2021) (alterations in original), aff'd, No. 21-4142, 2023 WL 4678683 (10th Cir. July 21, 2023); Ayala v. Am. Home Mortg. Servicing, Inc., 2011 WL 3319543, at *3 (D. Utah June 8, 2011) (same), report and recommendation adopted, 2011 WL 3319726 (D. Utah Aug. 1, 2011). Meta fails to explain how a mortgage loan is comparable for these purposes to Meta's relationships with its users.

72.   The States allege claims of unfair and/or deceptive acts or practices under the laws of thirteen states, namely Connecticut, Georgia, Hawai'i, Illinois, Kentucky, Louisiana, Maine, Michigan, Missouri, Nebraska, Pennsylvania, Rhode Island, and South Carolina. The personal injury plaintiffs additionally bring claims under the laws of the remaining six states: Alabama, Idaho, Massachusetts, Tennessee, Texas, and Vermont.

73.   To the extent other statutory arguments not raised by the parties exist, the Court does not here independently investigate and rule.

74.   Ala. Code § 8-19-5; Conn. Gen. Stat. Ann. § 42-110b(a); Haw. Rev. Stat. § 480-2(a); Idaho Code Ann. § 48-603; 815 Ill. Comp. Stat. Ann. 505/2; Ky. Rev. Stat. § 367.170; La. Stat. Ann. § 51:1405(A); Me. Rev. Stat. tit. 5, § 207; Mass. Gen. Laws Ann. ch. 93A, § 2(a); Mich. Comp. Laws Ann. § 445.903(1); Mo. Ann. Stat. § 407.010(7); Neb. Rev. Stat. Ann. § 59-1602; 73 Pa. Stat. Ann. § 201-3(a); 6 R.I. Gen. Laws Ann. § 6-13.1-2; S.C. Code Ann. § 39-5-20(a); Tenn. Code Ann. § 47-18-104(a) (“affecting the conduct of any trade or commerce” (emphasis supplied)); Tex. Bus. & Com. Code Ann. § 17.46(a); see also Ga. Code. Ann. § 10-1-393(a) (covering “acts or practices in the conduct of consumer transactions and consumer acts or practices in trade or commerce” (emphasis supplied)). Vermont instead prohibits “unfair or deceptive acts or practices in commerce,” without reference to trade. See Vt. Stat. Ann. tit. 9, § 2453(a).

75.   Ala. Code § 8-19-3(14); 815 Ill. Comp. Stat. Ann. 505/1(f); Ky. Rev. Stat. Ann. § 367.110; La. Stat. Ann. § 51:1402(10)(a); Me. Rev. Stat. tit. 5, § 206; Mass. Gen. Laws Ann. ch. 93A, § 1(b); Mich. Comp. Laws Ann. § 445.902(1)(g); 73 Pa. Stat. Ann. § 201-2(3); 6 R.I. Gen. Laws Ann. § 6-13.1-1(5); S.C. Code Ann. § 39-5-10(b); Tenn. Code Ann. § 47-18-103(24); Tex. Bus. & Com. Code Ann. § 17.45(6).Idaho defines “trade” and “commerce” to “mean the advertising, offering for sale, selling, leasing, renting, collecting debts arising out of the sale or lease of goods or services or distributing goods or services, either to or from locations within the state of Idaho, or directly or indirectly affecting the people of this state.” Idaho Code Ann. § 48-602(2).Hawai'i does not provide a statutory definition of “trade or commerce.” According to the Supreme Court of Hawai'i, conduct occurs in “trade or commerce” when it pertains to transactions “occur[ring] within a ‘business context.’ ” Cieri v. Leticia Query Realty, Inc., 80 Haw. 54, 65 (Haw. 1995); see also Baham v. Ass'n of Apartment Owners of Opua Hale Patio Homes, 2014 WL 2761744, at *19 (June 18, 2014). Whether a transaction occurs in a “business context” “must be determined on a case-by-case basis by an analysis of the transaction.” Cieri, 80 Haw. at 65; see also Dalesandro v. Longs Drug Stores California, Inc., 383 F. Supp. 2d 1244, 1250 (D. Haw. 2005) (noting the Cieri court “cited with approval” a six-factor text to determine “business context” used by the Supreme Judicial Court of Massachusetts). Meta does not advance any Hawai'i-specific arguments, and the Court considers its analysis below comprehensive of Hawai'i caselaw at this stage nonetheless.The Nebraska Consumer Protection Act provides that “[t]rade and commerce shall mean the sale of assets or services and any commerce directly or indirectly affecting the people of the State of Nebraska,” and “[a]ssets shall mean any property, tangible or intangible, real, personal, or mixed, and wherever situated, and any other thing of value.” Neb. Rev. Stat. Ann. § 59-1601(2)–(3).“Although the [Vermont Consumer Protection Act] does not define ‘in commerce,’ the Vermont Supreme Court has held that the ‘in commerce’ requirement narrows the VCPA's application to prohibit ‘only unfair or deceptive acts or practices that occur in the consumer marketplace.’ ” Hoehl Fam. Found. v. Roberts, 2020 WL 10790035, at *16 (D. Vt. Oct. 15, 2020) (quoting Foti Fuels, Inc. v. Kurrle Corp., 90 A.3d 885, 892 (Vt. 2013)).

76.   See, e.g., Stevens v. Motorists Mut. Ins. Co., 759 S.W.2d 819, 821 (Ky. 1988) (explaining that Kentucky legislature created the Act “in order to give Kentucky consumers the broadest possible protection for allegedly illegal acts” and Kentucky law requires that the statute be “liberally construed”);Price v. Long Realty, Inc., 502 N.W.2d 337, 342 (Mich. App. Ct. 1993) (“[B]ecause the MCPA is a remedial statute designed to prohibit unfair practices in trade or commerce, it must be liberally construed to achieve its intended goals.”);Aquilina v. Certain Underwriters at Lloyd's, 465 F. Supp. 3d 1088, 1099 (D. Haw. 2020) (“HRS § 480-2 ․ was constructed in broad language in order to constitute a flexible tool to stop and prevent fraudulent, unfair or deceptive business practices for the protection of both honest consumers and honest business [persons].” (alterations in original) (citation omitted));Garcia v. Foot Locker Retail, Inc., 293 A.3d 262, 266 (Pa. Super. Ct. 2023) (“These remedial statutes are all predicated on a legislative recognition of the unequal bargaining power of opposing forces in the marketplace” and shall “be construed liberally to effectuate that goal.” (citations omitted));Laccinole v. Rocket Mortg., LLC, 609 F. Supp. 3d 68, 74 (D.R.I. 2022);Taylor v. Medenica, 479 S.E.2d 35, 44 (S.C. 1996) (“The provision of any service constitutes commerce within the meaning of the” Act.);Fayne v. Vincent, 301 S.W.3d 162, 172 (Tenn. 2009) (“The Act is to be liberally construed in order to enable it to protect the consumer and to promote the other policies which motivated its passage.”).

77.   For example, in Georgia, an act or practice is not in “trade and commerce” unless it can “be said to have ‘impact’ on the consumer marketplace.” Henderson v. Gandy, 608 S.E. 2d 248, 252 (Ga. Ct. App. 2005) (quoting Zeeman v. Black, 273 S.E.2d 910, 915 (Ga. Ct. App. 1980)), aff'd 623 S.E.2d 465 (Ga. 2005); see also id. (at-issue transaction must occur “within the public consumer marketplace”).In Michigan, while “[t]here is no requirement that consumer goods be sold or purchased in order to constitute ‘trade and commerce,’ ” Florists' Transworld Delivery, Inc. v. Fleurop-Interflora, 261 F. Supp. 2d 837, 849 (E.D. Mich. 2003) (citation omitted), courts will consider whether a defendant's conduct “occurred as part of an effort to further the sale or lease of goods or services to consumers,” DIRECTV, Inc. v. Shouldice, 2003 WL 23200255, at *4 (W.D. Mich. Oct. 20, 2003); see also City of Charleston, S.C. v. Hotels.com, LP, 520 F. Supp. 2d 757, 776 (D.S.C. 2007) (holding as “plainly trade or commercial activity” where “[d]efendants provided consumers with hotel rooms in exchange for monetary consideration”).In Vermont, to “be considered ‘in commerce,’ the transaction must take place ‘in the context of [an] ongoing business in which the defendant holds himself out to the public.’ ” Foti Fuels, Inc., 90 A.3d at 892–93 (alterations in original) (quoting Zeeman v. Black, 273 S.E.2d 910, 915 (Ga. 1980)); see also Binette v. Dyer Libr. Ass'n, 688 A.2d 898, 907 (Me. 1996) (“the conduct of trade or commerce” covers “transaction[s] done in the ordinary course of a business or occupation”).

78.   For instance, certain governmental obligations or regulatory functions have been excluded from the meaning of “trade and commerce.” See L. Indus., LLC v. Dep't of Educ., 378 So. 3d 3, 8 (La. 2024) (state's conduct “in furtherance of its governmental function of providing safe and appropriate educational facilities” did not constitute engagement in “trade or commerce”); Garcia v. Foot Locker Retail, Inc., 293 A.3d 262, 270 (Pa. Super. Ct. 2023) (“[C]arrying out of a public duty, in this case the collection of sales tax, is not trade or commerce ․”); Health Promotion Specialists, LLC v. S.C. Bd. of Dentistry, 743 S.E.2d 808, 816 (S.C. 2013) (regulatory authority's “promulgation of a regulation” did not fall within “trade or commerce” “as it did not involve advertisement, sale, or distribution of services or property within a business context”).

79.   Separately, the States have presented some allegations that certain users pay for premium services on Meta's platforms. Even if a state has further requirements that the conduct must relate to a monetary exchange between the platform and its users—not apparent on this Court's review—the States' claims would survive at least to the extent premised on paying users.

80.   Alabama provides that “[a]ny person who commits one or more of the acts or practices declared unlawful under this chapter and thereby causes monetary damage to a consumer ․ shall be liable to each consumer ․” Ala. Code § 8-19-10(a). Texas permits a “consumer” to “maintain an action” against a defendant whose use of deceptive or unconscionable acts or practices “constitute a producing cause of economic damages or damages for mental anguish.” Tex. Bus. & Com. Code Ann. § 17.50(a). Vermont likewise provides that “[a]ny consumer who contracts for goods or services” and “sustains damages” as a result of prohibited unfair and deceptive acts and practices “may sue for appropriate equitable relief and may sue and recover from the seller, solicitor, or other violator the amount of his or her damages, or the consideration or the value of the consideration given by the consumer, reasonable attorney's fees, and exemplary damages ․” Vt. Stat. Ann. tit. 9, § 2461(b).

81.   See Ziencik v. Snap, Inc., No. 21-cv-7292, 2024 U.S. Dist. LEXIS 12105, *16 (C.D. Cal. Jan. 19, 2024) (“Plaintiffs have cited no authority for the idea that exchanging personally identifiable information for use of a free social networking application constitutes a ‘purchase’ or ‘lease’ under the CLRA ․”); In re Yahoo! Inc. Customer Data Sec. Breach Litig., 2017 WL 3727318, at *32 (N.D. Cal. Aug. 30, 2017) (“The mere fact that Yahoo gained some profit from Plaintiffs' use of Yahoo's free email services does not by itself show that Plaintiffs ‘purchased’ those services from Defendants” under the CLRA.); Claridge v. RockYou, Inc., 785 F. Supp. 2d 855 (N.D. Cal. 2011) (holding that plaintiff's “transfer of PII information to defendant in exchange for free applications” did not “constitute[ ] a ‘purchase’ or ‘lease’ under the CLRA”); see also Cal. Civ. Code § 1761(d) (“ ‘Consumer’ means an individual who seeks or acquires, by purchase or lease, any goods or services for personal, family, or household purposes.”).

82.   While Meta apparently intends for its motion to cover every state in the AG coalition, its motion only cites the laws of Arizona, California, Colorado, Delaware, Georgia, Illinois, Indiana, Kentucky, Michigan, Minnesota, Missouri, Nebraska, New Jersey, New York, North Dakota, Pennsylvania, Rhode Island, South Carolina, Virginia, Washington, and Wisconsin.

83.   Illinois's consumer-protection statute “expressly authorizes the Attorney General to enjoin illegal practices and to collect actual damages.” People ex rel. Hartigan v. Lann, 587 N.E.2d 521, 524 (Ill. App. Ct. 1992). While the statute explicitly authorizes restitution and not disgorgement, see 815 Ill. Comp. Stat. Ann. 505/7(a), Illinois courts recognize that restitution “flows from the basic policy that those who engage in proscribed conduct or practices surrender all profits flowing therefrom.” Lann, 587 N.E.2d at 524. As such, Illinois courts have permitted disgorgement as a remedy under its consumer-protection statute. See, e.g., People by Madigan v. CMK Invs., Inc., 2015 WL 4038896, at *2 (N.D. Ill. June 30, 2015) (striking damages-based affirmative defenses because “any monetary recovery by plaintiff is more in the nature of disgorgement than damages,” as authorized under ICFA); People, ex rel. Alvarez v. Rd. Am. Auto., Inc., 2014 IL App (1st) 120825-U, ¶ 22 (affirming trial judgment under ICFA which consisted of “$59,926.16 in Restitution; $29,208.30 in State[s] Attorney's investigative costs; $2,200,000.00 in Civil Penalties * * *; and $25,167.50 to be disgorged and paid into [a] cy pres account for consumer education purposes” (alterations in original)).Kentucky caselaw provides that its courts, in other contexts, have embraced the broader sense of restitution that encompasses unjust enrichment and disgorgement. See Haeberle v. St. Paul Fire & Marine Ins. Co., 769 S.W.2d 64, 67 (Ky. Ct. App. 1989) (The doctrine of unjust enrichment, “an equitable one, is applicable as a basis of restitution to prevent one person from keeping money or benefits belonging to another.”); Superior Steel, Inc. v. Ascent at Roebling's Bridge, LLC, 540 S.W.3d 770, 781 (Ky. 2017) (“[T]hese appellate court decisions” are “representative of Kentucky courts' willingness to adopt equitable measures where appropriate and wholly consistent with the principles in the Restatement (Third) of Restitution and Unjust Enrichment ․”).

84.   For example, the Minnesota Attorney General's parens patriae authority confers “the power to seek equitable restitution” when acting “on behalf of all Minnesotans harmed by a pattern and practice of fraudulent conduct.” State v. Minnesota Sch. of Bus., Inc., 935 N.W.2d 124, 133 (Minn. 2019). Subdivision 8.31(3) of the Minnesota Statutes “broadly authorizes the Attorney General to seek equitable relief to stop conduct that harms consumers.” Id. at 134. Subdivision 8.31(3)'s authorization extends to the scope of cases described in subdivision 8.31(1), which covers “violations of the law of this state respecting unfair, discriminatory, and other unlawful practices in business, commerce, or trade, and specifically, but not exclusively,” a set of ten laws enumerated in the subdivision. Minn. Stat. Ann. § 8.31(1). “The list of 10 specific laws following the broad grant of authority is not exclusive.” Findling v. Grp. Health Plan, Inc., 998 N.W.2d 1, 6–7 (Minn. 2023). In Findling, the Supreme Court of Minnesota held that the provision of the Minnesota Health Records Act (“MHRA”) “governing disclosure of, and patient access to, healthcare records” “is a law respecting unfair, discriminatory, and other unlawful practices in business, commerce, or trade,” even though the MHRA is not specifically enumerated in Subdivision 8.31(1). Id. at 7–8.

85.   In opposition, many states simply cited alternate provisions of their consumer-protection statutes, while other states provided some caselaw to indicate how appellate courts of their state might be inclined to approach this question. Dkt. No. 701-2 at 45–50. Because the burden here falls to Meta, the Court declines to limit the scope of available remedy at this stage.

86.   Meta argues on reply that the theory of restitution as described in Brown v. Google is not presented in the complaint. While there is some merit to this contention, complete theories of recovery need not be articulated in a complaint.

87.   Defendants cite to Edlow v. RBW, LLC, in which the First Circuit dismissed the plaintiff's deceptive acts and practices claim under Massachusetts law because “no reasonable fact-finder could conclude that [plaintiff's] reliance upon the allegedly false statements was reasonable.” 688 F.3d 26, 39 (1st Cir. 2012). In so determining, Edlow relied on (i) Aspinall's description of “interpretations of the term ‘deceptive’ under Federal law,” Aspinall, 813 N.E.2d at 488; and (ii) a District of Massachusetts opinion which held that a “plaintiff cannot prevail on” fraud, negligent misrepresentation, and unfair and deceptive trade practices “claims unless it can prove that its reliance on the allegedly false statements was reasonable,” Massachusetts Laborers' Health & Welfare Fund v. Philip Morris, Inc., 62 F. Supp. 2d 236, 241–42 (D. Mass. 1999). In light of Aspinall's explicit statements that reliance is not required to state a claim of deceptive acts and consistent authority upholding that proposition, the Court declines to dismiss Massachusetts claims for a failure to plead actual reliance absent contravening authority from the Supreme Judicial Court of Massachusetts.

88.   While certain states dispute whether Rule 9 applies to deceptive and acts and practices claims, see supra n.40, personal injury plaintiffs do not dispute Rule 9's application and instead focus on the standard applicable to claims of omissions.

89.   For example, under California's UCL, actual reliance “requires a plaintiff to allege that ‘the defendant's misrepresentation or nondisclosure was an immediate cause of the plaintiff's injury-producing conduct.’ ‘A plaintiff may ․ [do so] by showing that in its absence the plaintiff in all reasonable probability would not have engaged in the injury-producing conduct.’ ” Perkins v. LinkedIn Corp., 53 F. Supp. 3d 1190, 1220 (N.D. Cal. 2014) (quoting In re Tobacco II Cases, 207 P.3d 20, 39 (Cal. 2009)) (citations omitted). It need not be “the sole or even the predominant or decisive factor influencing [plaintiff's] conduct, only that it played a substantial part in the plaintiff's decision-making.” In re ZF-TRW Airbag Control Units Prod. Liab. Litig., 601 F. Supp. 3d at 766 (cleaned up) (citations omitted). “That one would have behaved differently can be presumed, or at least inferred, when the omission is material.” Madani v. Volkswagen Grp. of Am., Inc., 2019 WL 3753433, at *10 (N.D. Cal. Aug. 8, 2019) (quoting Daniel v. Ford Motor Co., 806 F.3d 1217, 1225 (9th Cir. 2015)).

90.   In Pearson, plaintiffs “urge[d] that a UTPA action based on failure to disclose does not require reliance because it would be ‘artificial’ to require a plaintiff to prove that it relied on undisclosed information. In response, defendant characterize[d] this as a ‘half-truth’ case, one that involves an alleged affirmative misrepresentation coupled with the alleged failure to disclose, and therefore one that requires proof of reliance.” 361 P.3d 3 at 28 (citation omitted). The court found that “[t]hose arguments miss the mark” and “decline[d] the parties' mutual invitations to reduce the analysis to an exercise in attaching labels․ The answer requires reasoned analysis of the claim, not labeling.” Id. (citation omitted). Plaintiffs' “hybrid” framing of defendant's representation—“part affirmative misrepresentation and part failure-to-disclose”—“did not change the causative link that their” theory “depended on. Under that theory, proof of reliance on the alleged misrepresentation was integral to the plaintiffs’ ” claim. Id.

91.   Those that do, defendants note, only make conclusory assertions of reliance. E.g., Baker SFC (No. 23-cv-01578) (“Had the omitted information been disclosed, the injuries that Plaintiff suffered would have been avoidable and avoided.”).

92.   Note also that Count 7 of the 2AMC is unambiguous in seeking to impose liability on defendants' affirmative statements as well as omissions. (See 2AMC ¶ 964 (“Each Defendant violated consumer protection laws, through its use of false and misleading misrepresentations or omissions of material fact relating to the safety of its respective social media products.”); id. ¶ 965 (“Each Defendant uniformly communicated the purported benefits and safety of using its respective social media products while failing to disclose the serious harms related to use, particularly to youth. Defendants made these representations to the public, the government, and consumers.”).)

93.   Meta claims that a showing of ascertainable loss is required under the laws of Alabama, California, Connecticut, Delaware, Idaho, Iowa, Kentucky, Louisiana, Maine, Missouri, New Jersey, Ohio, Oregon, Pennsylvania, Rhode Island, South Carolina, Virginia, and West Virginia. (Dkt. No. 517 at 63 n.69.) Plaintiffs have not disputed that these states require a showing of ascertainable loss for a private plaintiff to bring suit.

94.   Notably, plaintiffs allege as above COPPA violations against all defendants via state consumer-protection claims. Neither defendants in their joinder nor plaintiffs in their opposition address COPPA, and for that matter the 2AMC appears bereft of any specific COPPA allegations. The Court declines to make any specific ruling as to COPPA under private plaintiffs' state consumer-protection claims absent briefing.

95.   At best, plaintiffs' opposition provides a string citation to a handful of allegations in the 2AMC. (Dkt. No. 600 at 7 (citing 2AMC ¶¶ 96–124, 167–70, 184–85, 366, 369, 390–91, 540, 682, 802–03).) As regards affirmative statements by the non-Meta defendants, none are additive. Most concern the general knowledge that the defendants' product designs cause harm, and some are specific to Meta. (2AMC ¶¶ 96–124, 167–70, 184–85, 366, 369, 390–91.) One allegation as to Snap is described above. (Id. ¶ 540.) Another references TikTok's “funny, cheerful ads featuring smiling families and funny images.” (Id. ¶ 682.) The remaining allegation concerns Google's launch of an AI in 2018 which it claimed “drastically improved” CSAM detection distributed on YouTube, and Google's general claim that it will continue to invest in technology and organizations to fight the dissemination of CSAM. (Id. ¶¶ 802–03.)

96.   As previously noted, Georgia courts do not appear to recognize negligent misrepresentation by omission. See Ali v. Fleet Fin., Inc. of Georgia, 500 S.E.2d 914, 915 (Ga. Ct. App. 1998) (“[B]ecause Fleet did not make any representations, Ali cannot set forth the requisite elements of negligent misrepresentation.”); Intellicig USA, LLC v. CN Creative Ltd., No. 15-cv-1832, 2017 WL 11634374, at *7 n.6 (N.D. Ga. Mar. 6, 2017) (“[C]ases strongly suggest[ ] that a negligent misrepresentation claim requires an affirmative representation.”). Further, plaintiffs have conceded as much. (Dkt. No. 661 at 1 n.†, 2).As previously noted, Maryland has, at best, not recognized “negligent concealment,” Chassels v. Krepps, 174 A.3d 896, 902 (Md. Ct. Spec. App. 2017) (negligent concealment “has [not] been recognized as a tort in Maryland” (quoting Shaw v. Brown & Williamson Tobacco Corp., 973 F. Supp. 539, 550 n.2 (D. Md. 1997)) (alteration in original)), and the Court declines to expand Maryland law.As to North Carolina, plaintiffs rely on a case which permitted a claim of negligent misrepresentation to proceed when based off of “a series of false and fraudulent representations, omitted material facts and told deceptive half-truths.” Bradshaw v. Maiden, 2015 WL 4720387, at *2 (N.C. Super. Ct. Aug. 10, 2015), aff'd, 881 S.E.2d 643 (N.C. Ct. App. 2022), aff'd, ordered not precedential, 898 S.E.2d 268 (N.C. 2024). Yet, Bradshaw describes a claim based on affirmative misrepresentations rendered deceptive by material omissions; that is, Bradshaw does not support recognizing a claim for negligent misrepresentation based on omissions alone. Defendants provide clear authority to the contrary that “negligent omissions ․ as opposed to negligent misrepresentations cannot form the basis of a claim for negligent misrepresentation under North Carolina law.” Bonham v. Wolf Creek Acad., 767 F. Supp. 2d 558, 570 (W.D.N.C. 2011); see also DeGorter v. Capitol Wealth, Inc., 2016 WL 3944086, at *10 n.2 (N.C. Super. May 31, 2016).In Ohio, “Negligent misrepresentation does not lie for omissions; there must be some affirmative false statement.” Textron Fin. Corp. v. Nationwide Mut. Ins. Co., 684 N.E.2d 1261, 1269 (Ohio Ct. App. 1996).

97.   “[A] cause of action for misrepresentation requires an affirmative statement, not an implied assertion.” RSB Vineyards, LLC v. Orsi (2017) 15 Cal. App. 5th 1089, 1102; Republic Bank & Tr. Co. v. Bear, Stearns & Co., 707 F. Supp. 2d 702, 714 (W.D. Ky. 2010) (Negligent misrepresentation “requires an affirmative false statement; a mere omission will not do.”), aff'd, 683 F.3d 239 (6th Cir. 2012).As to Washington, the parties submit opposing statements from the Supreme Court of Washington sitting en banc. Compare Ross v. Kirner, 172 P.3d 701, 704 (Wash. 2007) (en banc) (“An omission alone cannot constitute negligent misrepresentation, since the plaintiff must justifiably rely on a misrepresentation.”), with Van Dinter v. Orr, 138 P.3d 608, 610 (Wash. 2006) (en banc) (“If a party has a duty to disclose information, the failure to do so can constitute negligent misrepresentation.”). Ross being the more recent, and direct, decision is controlling. Van Dinter can be reconciled therewith but such analysis would require more than is needed here.

98.   On a motion for reconsideration, the district court in Burket later vacated a portion of its order on an unrelated issue. See No. 05-72110, 2008 WL 2478308 (E.D. Mich. June 17, 2008).

99.   This Court previously recognized Texas as a state which permits claims of negligent misrepresentation by omission. See In re Soc. Media Adolescent Addiction/Pers. Inj. Prod. Liab. Litig., 2024 WL 1786290, at *6 & n.11 (citing Brown & Brown of Texas, Inc. v. Omni Metals, Inc., 317 S.W.3d 361, 384 (Tex. App. 2010) (“In the prior appeal of this case, the Fourteenth Court of Appeals pointed out that a misrepresentation need not be an affirmative misrepresentation of fact.”)); Baker v. Great N. Energy, Inc., 64 F. Supp. 3d 965, 979 (N.D. Tex. 2014) (“[T]o the extent [plaintiff's] negligent misrepresentation claim is based on Defendants' alleged omissions, these assertions also fail, because ‘in Texas, non-disclosures cannot be negligent unless there is a duty to disclose’ ․” (quoting Coburn Supply Co. v. Kohler Co., 342 F.3d 372, 377 (5th Cir. 2003))). The Court acknowledges that, despite this language from Brown and Baker, at least one Texas court has signaled that Texas has not decided whether a claim of negligent misrepresentation can be based on omissions alone. See Lindsey Constr., Inc. v. AutoNation Fin. Servs., LLC, 541 S.W.3d 355, 366 (Tex. App. 2017) (“We presume, without deciding, that a negligent-misrepresentation claim may be based on a failure to disclose rather than an affirmative misrepresentation.”). The District Court of New Jersey has determined that “an affirmative representation is required to assert a negligent misrepresentation claim under Texas” law. Tijerina v. Volkswagen Grp. of Am., Inc., 2023 WL 6890996, at *35 & n.22 (D.N.J. Oct. 19, 2023). Given the indeterminate caselaw, the Court declines to dismiss claims of negligent misrepresentation by omission under Texas law at this stage.

100.   For Hawai'i, see Newcomb v. Cambridge Home Loans, Inc., 861 F. Supp. 2d 1153, 1163 (D. Haw. 2012) (“Plaintiff does not cite, nor has the Court found, any authority recognizing a claim for negligent concealment.”).For Kansas, see Plastic Packaging Corp. v. Sun Chem. Corp., 136 F. Supp. 2d 1201, 1206–07 & n.2 (D. Kan. 2001) (noting that “Kansas courts have not recognized a cause of action for negligent omission” and dismissing the claim for failure to allege a duty to disclose), and Hanson v. Hackman Corp., 192 P.3d 1130, at *6 (Kan. Ct. App. 2008) (“Kansas has not adopted the tort of negligent nondisclosure, although neither has the tort directly been rejected.”).For Oregon, see Vigilante.com, Inc. v. Argus Test.com, Inc., No. 04-cv-413, 2005 WL 2218405, at *16 (D. Or. Sept. 6, 2005) (negligent misrepresentation claim requires showing of “plaintiff's reasonable reliance on ․ false representations or omissions”). Defendants note that neither of the authorities Vigilante.com relied upon mentioned omissions. See Conway v. Pac. Univ., 924 P.2d 818 (Or. 1996) (en banc); Prosser v. Safeco Ins. Co., 2001 U.S. Dist. LEXIS 8341 (D. Or. June 15, 2001).For South Carolina, plaintiffs cite Pruitt v. Morrow, which Meta notes only recognized “actions based upon negligent or reckless nondisclosure of land defects.” 342 S.E.2d 400, 401 (S.C. 1986). Meta provides no alternate citation for South Carolina.For Virginia, as this Court has previously noted, compare Guerin v. Mechanicsville Properties, L.L.C., 81 Va. Cir. 184 (Va. Cir. Ct. 2010) (“[C]oncealment can give rise to constructive fraud only in cases in which there is a duty to disclose the concealed fact.”), with N. Virginia Eye Inst., P.C. v. Cynosure, LLC, No. 21-cv-00008, 2021 WL 1554887, at *4 (W.D. Va. Apr. 20, 2021) (“[C]onstructive fraud by concealment is not a viable claim under Virginia law ․”).For Wisconsin, as this Court has previously noted, compare Ramsden v. Farm Credit Servs. of N. Cent. Wisconsin ACA, 590 N.W.2d 1, 5 & n.4 (Wis. Ct. App. 1998) (“[N]egligent failure to disclose may give rise to a claim for relief, if there is a duty to speak.”), with Amcore Bank v. Heus Mfg. LLC, 812 N.W.2d 540, ¶ 22 (Wis. Ct. App. 2012) (“It remains an open question in Wisconsin, however, whether a failure to disclose can support a claim for negligent misrepresentation.”).

101.   Meta also moves to dismiss the Florida AG's complaint, filed in the U.S. District Court for the Middle District of Florida, for improper venue and lack of personal jurisdiction. At the April 19, 2024 hearing on Meta's motion, the Court granted the State leave to replead its claims with respect to jurisdiction. Dkt. No. 785, Tr. of April 19, 2024 Case Management Conference at 126:23–127:2. The Court new motion is fully briefed. As such, Meta's motion to dismiss the Florida AG's original complaint for lack of improper venue or personal jurisdiction is DENIED as moot.

YVONNE GONZALEZ ROGERS UNITED STATES DISTRICT COURT JUDGE