D.C. v. Meta: Complaint for Violations of the Consumer Protection Procedures Act

Raw text

Transcribed text as recorded, including headers and formatting fragments.

# IN THE SUPERIOR COURT OF THE DISTRICT OF COLUMBIA CIVIL DIVISION \| DISTRICT OF COLUMBIA, Plaintiff, v. V. \| Case No.:2023-CAB-006550 Judge:Hon. Neal E.Kravitz Next Event:Initial Hearing Date:May 24,2024 \| \| \-\-\- \| \-\-\- \| \| META PLATFORMS,INC.,INSTAGRAM,LLC, Defendants. \| \| ## PRAECIPE TO FILE UNSEALED COMPLAINT Pursuant to this Court's November 13,2023 Order, which instructed Plaintiff District of Columbia (the "District") to file a clean, unredacted version of the Complaint on the public record,the District hereby attaches the unsealed Complaint as Exhibit A. Dated: November 14,2023 Respectfully submitted, BRIAN L. SCHWALB Attorney General for the District of Columbia JENNIFER C. JONES /s/ Adam R. Teitelbaum ADAM R. TEITELBAUM (#1015715) Director, Office of Consumer Protection KEVIN VERMILLION (#1739318) Deputy Director, Office of Consumer Protection * * * JORGE A. BONILLA LOPEZ (#1632867) Assistant Attorney General LYDIA E. MENDEZ (#90005211) Assistant Attorney General Office of the Attorney General for the District of Columbia 400 6th Street NW, 10th Floor Washington, D.C. 20001 Tel: (202) 741-5226 [adam.teitelbaum@dc.gov](mailto:adam.teitelbaum@dc.gov) [kevin.vermillion@dc.gov](mailto:kevin.vermillion@dc.gov) [jorge.bonillalopez@dc.gov](mailto:jorge.bonillalopez@dc.gov) [lydia.mendez@dc.gov](mailto:lydia.mendez@dc.gov) /s/ Jimmy R. Rock Jimmy R. Rock [jrock@edelson.com](mailto:jrock@edelson.com) EDELSON PC 1255 Union Street NE, 7th Floor Washington, DC 20002 Tel: (202) 987-6302 Theo Benjamin\* [tbenjamin@edelson.com](mailto:tbenjamin@edelson.com) Emily Penkowski Perez\* [epenkowski@edelson.com](mailto:epenkowski@edelson.com) EDELSON PC 350 North LaSalle Street, Suite 1400 Chicago, Illinois 60654 Tel: (312) 589-6370 Attorneys for the District of Columbia \*Pro Hac Vice applications forthcoming * * * ## CERTIFICATE OF SERVICE I certify that on November 14, 2023, a copy of the foregoing Praecipe to File Unsealed Complaint was filed and served electronically via eFileDC on all counsel of record. /s/ Jimmy R. Rock Attorney for the District of Columbia * * * ## EXHIBIT A * * * # IN THE SUPERIOR COURT OF THE DISTRICT OF COLUMBIA Civil Division | DISTRICT OF COLUMBIAa municipal corporation400 Sixth Street,NWWashington,D.C.20001,Plaintiff,v.META PLATFORMS,INC.,INSTAGRAM,LLC,Serve on:META,INC.1 Hacker WayMenlo Park,CA94025,Defendants. | Case No.:2023-CAB-006550Judge:Hon.Neal E.KravitzJURY TRIAL DEMANDED | | --- | --- | ## COMPLAINT FOR VIOLATIONS OF THE CONSUMER PROTECTION PROCEDURES ACT The District of Columbia (the "District") by the Office of the Attorney General, brings this action against Defendants Meta Platforms, Inc. and Instagram, LLC (collectively "Meta") for violations of the District's Consumer Protection Procedures Act ("CPPA"), D.C.Code $ \\gg$ 283901, et seq. In support of its claims, the District states as follows: ## INTRODUCTION * * * Platforms and takes advantage of children—users under the age of eighteen—for its own financial gain. Meta’s strategic choice to exploit and manipulate children in this way, and to inflict upon them long-lasting harms to their well-being, is unlawful. 2. Because children are uniquely vulnerable to addictive technologies, Meta targets children and incentivizes its teams to increase engagement and time children spend on its Social Media Platforms to earn more money. Even though users can create accounts on the Platforms without paying Meta any money, the Platforms are not free. Meta collects users’ data and monopolizes their time, then marketers pay Meta to place their advertisements before the eyeballs of its billions of users. The more time users spend on the Platforms, the more marketers are willing to pay. 3. To maximize engagement, Meta has developed and continually refines deceptive and addictive features designed to trick users—particularly children—into spending more time on its Platforms. Aware that children’s developing brains are more susceptible to certain manipulative tactics, Meta exploits these neurological vulnerabilities through features including: (a) algorithms that manipulate users into staying on the app longer in search of a “hit” of dopamine; (b) audio, visual, and vibrating alerts that incessantly recall children to Meta’s Social Media Platforms while at school and during the night; and (c) content-presentation formats, such as “infinite scroll,” “autoplay,” and short-form videos known as “Reels,” designed to discourage children’s attempts to self-regulate and disengage with Meta’s products. 4. There are strong and well-researched links between young peoples’ excessive use of social media and negative outcomes, including depression, anxiety, insomnia, and interference with education and daily life. Meta has known of these harms, despite its contrary public statements. Meta’s own internal research documents show its awareness that its products harm children. Indeed, internal studies that Meta commissioned (and kept private until they were leaked by a whistleblower and reported by The Wall Street Journal) reveal that Meta has known for years about these serious harms to children from use of its Social Media Platforms. 5. Meta’s Social Media Platforms are built to be irresistible even in the face of teenagers’ struggles: Meta’s internal surveys show it makes “teens feel addicted” to the extent that, “like addicts, they feel that they are unable to stop themselves from being on” the Platforms. The primary mechanism for this addiction involves dopamine: Instagram delivers teens “a dopamine hit,” knowing that teen brains are “much more sensitive” than adult brains to the “insatiable” need for “feel good dopamine effects.” 6. Meta’s tactics have worked. In 2022, 62% of teens reported using Instagram, and 32% reported using Facebook.$^1$ Almost half of teens reported checking Instagram at least once a day, with 27% reporting checking it several times a day and 10% reporting checking it almost constantly.$^2$ Data has also shown that middle school- and high school-aged children are averaging 3.5 hours a day on social media—with nearly 25% spending five or more hours a day on social media.$^3$ 7. Screentime is a particular problem in the District. According to the CDC’s Youth Risk Behavioral Survey (“YRBS”), nearly 70% of the District’s high school students spend on average more than three hours a school day on screen time (not counting schoolwork) and just a $^1$ Emily A. Vogels and Risa Gelles-Watnick, Teens and social media: Key findings from Pew Research Center surveys, PEW RSCH. CTR. (Apr. 24, 2023), [https://www.pewresearch.org/shortreads/2023/04/24/teens-and-social-media-key-findings-from-pew-research-center-surveys/](https://www.pewresearch.org/shortreads/2023/04/24/teens-and-social-media-key-findings-from-pew-research-center-surveys/). $^2$ Id. $^3$ U.S. DEP’T OF HEALTH & HUM. SVCS., Social Media and Youth Mental Health: The U.S. Surgeon General’s Advisory 4 (2023), [https://www.hhs.gov/sites/default/files/sg-youth-mentalhealth-social-media-advisory.pdf](https://www.hhs.gov/sites/default/files/sg-youth-mentalhealth-social-media-advisory.pdf). * * * quarter get eight hours of sleep a night.4 This phenomenon extends to the District’s middle school children as well, whose screen time goes up and sleep goes down over time. In sixth grade, 62% of District students spend three or more hours on screen time and 54.4% get eight hours of sleep a night. By eighth grade, 72.2% of students spend three or more hours of screen time and only 41.1% sleep for the recommended eight hours. 8. An advisory issued by the U.S. Surgeon General in 2023 noted that “the current body of evidence indicates that while social media may have benefits for some children and adolescents, there are ample indicators that social media can also have a profound risk of harm to the mental health and well-being of children and adolescents.”5 9. Meta is helping to drive today’s teenagers into record-high levels of anxiety and depression. Data from the Centers for Disease Control and Prevention show that most adolescent girls reported feeling “persistently sad or hopeless” in 2021. Approximately one in four (24%) teenage girls reported that they had made a suicide plan, and forty percent (40%) of high school students described mental health challenges so dire that “they could not engage in their regular activities for at least two weeks during the previous year.”6 10. These numbers are particularly acute in the District, which has seen an alarming worsening of mental health trends among adolescents since 2007. In 2021, almost half (47.7%) 4 2021 District of Columbia Youth Risk Behavior Survey Results: District of Columbia (Including Charter Schools) High School Survey, OFF. OF THE STATE SUPERINTENDENT OF EDUC., [https://osse.dc.gov/sites/default/files/dc/sites/osse/page\_content/attachments/2021DCBH%20Summary%20Tables.pdf](https://osse.dc.gov/sites/default/files/dc/sites/osse/page_content/attachments/2021DCBH%20Summary%20Tables.pdf). 5 U.S. DEP’T OF HEALTH & HUM. SVCS., Social Media and Youth Mental Health: The U.S. Surgeon General’s Advisory 4 (2023), [https://www.hhs.gov/sites/default/files/sg-youth-mental-health-social-media-advisory.pdf](https://www.hhs.gov/sites/default/files/sg-youth-mental-health-social-media-advisory.pdf). 6 CTRS. FOR DISEASE CONTROL AND PREVENTION, Youth Behavior Risk Survey: Data Summary & Trends Report 2 (2023), [https://www.cdc.gov/healthyyouth/data/yrbs/pdf/YRBS\_Data-Summary-Trends\_Report2023\_508.pdf](https://www.cdc.gov/healthyyouth/data/yrbs/pdf/YRBS_Data-Summary-Trends_Report2023_508.pdf). * * * of the District’s high school girls self-reported episodes of psychological distress, including persistently feeling sad or hopeless. Similarly, self-reports of suicidal thoughts, plans, or attempts have increased among young people in the District since 2007 and remained consistently high. In 2021, over a quarter (28%) of all middle school students and over a third (36.9%) of all middle school girls reported seriously considering attempting suicide.$^7$ 11. The steep increase in the number of teens who report psychological distress, including suicidal thoughts, has coincided with the dramatic growth of social media use in the United States. See Figures 1 and 2 below. (Figure 1.)$^8$ $^7$ 2021 District of Columbia Youth Risk Behavior Survey Results: District of Columbia (Including Charter Schools) High School Survey 15, 16-19, OFF. OF THE STATE SUPERINTENDENT OF EDUC., [https://osse.dc.gov/sites/default/files/dc/sites/osse/page\_content/attachments/2021DCBH%20Summary%20Tables.pdf](https://osse.dc.gov/sites/default/files/dc/sites/osse/page_content/attachments/2021DCBH%20Summary%20Tables.pdf). $^8$ Katherine Schaeffer, U.S. has changed in key ways in the past decade, from tech use to demographics, PEW RSCH. CTR. (Dec. 20, 2019), [https://www.pewresearch.org/short-reads/2019/12/20/key-ways-us-changed-in-past-decade](https://www.pewresearch.org/short-reads/2019/12/20/key-ways-us-changed-in-past-decade). * * * 12. These addictive features and design elements of Meta’s Social Media Platforms are unfair to consumers. For that reason alone, Meta has violated the CPPA. 13. Separately, Meta has violated the CPPA by deceptively misrepresenting to consumers that its Social Media Platforms are safe for children and by deceptively concealing that its Social Media Platforms are manipulatively designed to promote children’s prolonged and unhealthy engagement. These misrepresentations work to facilitate the harms that the addictive features and design elements of Meta’s Social Media Platforms inflict on consumers, especially children. 9 Jean M. Twenge, Increases in Depression, Self-Harm, and Suicide Among U.S. Adolescents After 2012 and Links to Technology Use: Possible Mechanisms, 2:1 PSYCHIATRIC RES. AND CLINICAL PRACTICE 20 (Mar. 27, 2020), [https://prcp.psychiatryonline.org/doi/10.1176/appi.prcp.20190015](https://prcp.psychiatryonline.org/doi/10.1176/appi.prcp.20190015). * * * 14. Meta made, and continues to make, these misrepresentations knowing that they are untrue—knowing that its Social Media Platforms are, in fact, designed to ensnare children, that children who use the Platforms are likely to be exposed to harmful content, and that design features that Meta promotes, like offering plastic surgery camera filters and displaying “likes,” are particularly harmful to children. 15. As part of its deception, Meta routinely publishes profoundly misleading public reports that purported to show impressively low rates of negative and harmful experiences by users of its Social Media Platforms. At the same time, Meta has secretly maintained a parallel set of internal data showing shockingly high rates of harms experienced by its users. By publishing the favorable data from the Community Standards Enforcement Reports (“CSER”) while concealing the alarming data from the Bad Experiences & Encounters Framework (“BEEF”) and Tracking Reach of Integrity Problems Survey (“TRIPS”), both of which showed frequent user harms, Meta has misrepresented to consumers that its Social Media Platforms are far safer for children than they actually are. 16. By making deceptive statements and concealing material information about the alleged safety of its Social Media Platforms, Meta has done far more than publish harmful content. 17. Protecting the District’s children so that they live healthy, hopeful lives is at the core of the Attorney General’s statutorily established duty to uphold the public interest. In this case, Meta—one of the world’s largest and most well-resourced companies—has been exploiting and deceiving the District’s children for the sake of its own profits, in violation of District law. * * * 18. The Attorney General brings this enforcement action to safeguard the District’s children and put an end to Meta’s unfair and deceptive conduct. ## PARTIES 19. Plaintiff District of Columbia (the “District”) is a municipal corporation empowered to sue and be sued and is the local government for the territory constituting the permanent seat of the federal government. The District brings this case through the Attorney General for the District of Columbia, who is the chief legal officer for the District. The Attorney General is responsible for upholding the public interest and is also specifically authorized to enforce the District’s consumer protection laws, D.C. Code § 1-301.81(a)(1), including the CPPA. D.C. Code § 28-3909. 20. Defendant Meta Platforms, Inc. (“Meta Platforms”) is a Delaware corporation with its principal place of business at 1601 Willow Road, Menlo Park, CA 94025. As relevant here, Meta Platforms develops, markets, and operates social media platforms and other internet-based platforms and products, including Facebook and, through its wholly-owned subsidiary, Instagram. Meta Platforms was formerly known as Facebook, Inc. until it changed its corporate name in October 2021. 21. Facebook is a social network that allows users to share and view content including “statuses,” photographs, and videos; join groups; buy and sell products; and message others. 22. Defendant Instagram, LLC, offers a mobile application (“Instagram”) that enables users to share and view content such as photographs and videos online and over social networks. Instagram LLC is a limited liability company incorporated in Delaware with its principal place of business at 1601 Willow Road, Menlo Park, CA 94025. Meta Platforms purchased Instagram in * * * 2012 for over one billion dollars. Instagram is currently a wholly-owned subsidiary of Meta Platforms. Meta Platforms asserts complete control over Instagram and there is heavy overlap in the operations and personnel between Meta Platforms and Instagram. 23. Meta Platforms, acting on its own and through and/or in concert with other wholly-owned subsidiaries, engages in the unlawful conduct alleged herein. Meta Platforms is heavily involved in and has final say over the design and marketing of its Social Media Platforms operated by itself and its subsidiaries, including Instagram. Defendants are collectively referred to throughout this Complaint as “Meta.” ## JURISDICTION AND META’S CONTACTS WITH THE DISTRICT OF COLUMBIA 24. This Court has subject matter jurisdiction over the claims in this Complaint pursuant to D.C. Code §§ 11-921 and 28-3909. 25. This Court has personal jurisdiction over all Defendants pursuant to D.C. Code § 13-423(a). 26. Meta engages in trade practices, which include supplying “consumer goods” and “consumer services of all types,” D.C. Code § 28-3901, in the District. At all times material to this Complaint, acting alone or in concert with its subsidiaries, Meta has advertised, marketed, and distributed Meta’s Social Media Platforms to consumers throughout the District and makes a substantial profit selling District consumers’ user data and time to advertisers. 27. Meta is registered to do business in the District and has made repeated filings with the former District of Columbia Department of Consumer and Regulatory Affairs (“DCRA”) (now the Department of Licensing and Consumer Protection), registering as a foreign corporation in the District since at least 2009. In these filings with DCRA, Meta has admitted that it conducts business in the District of Columbia. * * * 28. Meta has also maintained a Washington, D.C. office for over thirteen years, since at least 2010. Meta’s current Washington, D.C. office is located at 575 7th Street NW and has hundreds of employees. 29. Meta enters into contracts with each of its District users to provide them with online social networking services, including through Facebook and Instagram. Meta provides District users with mobile applications, readily available on smartphone devices, throughout the entire District. 30. As of April 2023, Instagram alone has a total of 570,435 daily active users in the District and 920,581 monthly active users in the District—including 55,580 daily active teenage users and 94,656 monthly active teenage users. 31. Meta encourages children, including tens of thousands of teenagers in the District, to use its products by viewing, liking, commenting on, and uploading videos, photos, and posts on Meta’s Social Media Platforms. 32. Meta also specifically conducts business related to the safety of its Platforms in the District. The Meta office in Washington, D.C. includes teams related to “Advertising Technology,” “Communications and Public Policy,” and “Software Engineering.” Meta’s current job openings in the District relate to topics such as infrastructure monetization, consumer product marketing, and analytics and data science for Meta’s Social Media Platforms. 33. Meta also routinely holds meetings and summits in Washington, D.C. with company leadership, youth, and other stakeholders relating to topics such as youth safety and well-being on Instagram and Facebook, including meetings in at least 2018, 2019, and 2022. For example, in 2018, Meta conducted roundtable discussions regarding bullying with interfaith communities here in the District, and in March 2018, Meta held a Global Safety Network * * * Summit in Washington, D.C. In 2019, Meta held a meeting in Washington, D.C. with the Vice President & Executive Director of the National Center on Sexual Exploitation in the District relating to sex trafficking on Instagram, and in September 2019 Meta held a “Design Jam” session collaborating with youth and other stakeholders on “Transparency, Control, and Safety for Young People.” In December 2022, Meta held its first summit focused on youth safety and well-being online in Washington, D.C. 34. Meta is aware that it has users in the District, and thus that it has been misrepresenting to District consumers that its Social Media Platforms are safe for children and that their features are not manipulative or designed to promote children’s prolonged and unhealthy engagement. 35. Further, Meta executives, including CEO Mark Zuckerberg and Global Head of Safety Antigone Davis, have made misrepresentations of material fact regarding the safety of its Social Media Platforms in the District while testifying before Congress. FACTUAL ALLEGATIONS ## I. Meta’s business model seeks to increase advertising revenue by keeping children engaged with its Platforms for as much time as possible. 36. Meta’s core business model across all of its Social Media Platforms rests on monetizing users’ time and data through increasing engagement (a metric known as time spent) on its Platforms. Meta constantly strives to maximize user engagement on its Platforms so it can sell more targeted advertising opportunities. 37. Meta’s content personalization algorithms are designed solely with its profit-making in mind: to capture users’ attention and keep them engaged. These algorithms do not promote any specific message. Rather, the algorithms function on a user-by-user basis, operating to detect which types of content each individual is likely to engage with and then displaying more of that type of content continuously to maximize time spent (and user data collected). As a result, users often find themselves unwittingly and infinitely scrolling for hours. 38. The longer a user stays on Meta’s Platforms, the more ads the user sees. The effective delivery of targeted ads is a pivotal factor in Meta’s ability to generate revenue. Advertisers can target ads by selecting from a variety of highly targeted user demographics (such as age, gender, and location) collected by Meta. 39. When Meta succeeds in maintaining a user’s interest through its content personalization algorithms, Meta is able to collect more data and serve more targeted advertisements to the user. Meta can then also apply its personalization algorithms to track ads users engage with and tailor each user’s feed with more effective advertising. The result is a very lucrative business of turning users’ attention and data into sales of highly targeted advertising. 40. Only recently did Meta stop allowing advertisers to target teenage users by gender, interests, activity on the Platform, or precise location. However, it still utilizes the same strategies to collect data and keep children engaged through its personalization algorithm, and children continue to be targeted by advertisers and see ads on the Platforms, which generates revenue for Meta. 41. Meta is especially financially motivated to attract and retain children on its Social Media Platforms. Teenagers are one of the most valuable advertising demographics, so brands will look to advertise where the teenagers are. Converting children into lifelong users also ensures the future success of Meta’s Social Media Platforms. 42. When Facebook’s child userbase declined in the 2010s, Meta embarked on a plan to “win back” the teen market to Facebook. This included a concerted effort in the United States to emphasize the “social entertainment” market. Since at least 2015, Meta has focused on driving time spent by teens on its Social Media Platforms and focused its business model on attracting and retaining teen users. 43. Internally, Meta focuses on youth engagement in assessing the long-term success of its Social Media Platforms, carefully tracking each hour teens spend on its Platforms. 44. For example, in 2019, Meta set one of Instagram’s quarterly goals to hitting two million hours of teen watch time on IGTV (short for Instagram TV), which is Meta’s long-form video feature. 45. In another example, an internal email circulated in September 2018 revealed that Meta discusses its youngest users in terms of their financial value to the company: “The lifetime value \[defined by Meta in this same email chain as a metric of profit\] of a 13 y/o teen is roughly $270 per teen.” 46. Externally, Meta has deceptively denied that it places a monetary value on children. In 2021, Senator Amy Klobuchar asked Antigone Davis, Meta’s Global Head of Safety, what Meta believed the lifetime monetary value of children who use Meta products was; Davis responded “that’s just not the way we think about it.” Davis also denied that Meta “considered the profit value of developing products when \[Meta\] make\[s\] their decisions of how those products look,” testifying that this would be a “terrible business model.” 47. In 2013, shortly after it acquired Instagram, Meta introduced the ability to create “sponsored posts.” This transformed Instagram from a pure social network into an advertising platform: advertisers could pay to put their advertisements in front of Instagram users’ eyes. Instagram advertising soon became pervasive. 48. Instagram has become Meta’s most successful social media platform in attracting and retaining children, and Meta is intensely focused on retaining this young audience. * * * 49. Both Facebook and Instagram have consumed the time and attention of a large percentage of the population nationally and in the District specifically. In 2023, Facebook had 226,387,306 active monthly users in the United States, 68% of the population, with 10,322,887 active monthly users under eighteen. At the same time, Facebook had 601,145 monthly active users in the District, which is 89.5% of the District’s population. 50. Instagram is similarly popular, but with even higher levels of engagement among under-eighteen users. In 2023, Instagram had 192,073,171 active monthly users in the United States, over 58% of the population, with 32,389,032 users under eighteen. As of April 2023, Instagram had 570,435 daily active users in the District, accounting for 85% of the District’s total population, and 920,581 monthly active users. 51. In 2022, 62% of teens reported using Instagram.$^{10}$ This rate is higher for teen girls, with 69% reporting using Instagram.$^{11}$ Almost half of teens reported checking Instagram at least once a day, with 27% reporting checking it several times a day and 10% (still millions of users) reporting checking it almost constantly.$^{12}$ 52. The same carries through in the District. In April 2023, Instagram had 55,580 daily active teen users and 94,656 monthly active teen users in the District. When compared against the population of minors in the District, these numbers show striking rates of use among the District’s children. There are about 124,000 children in the District under the age of eighteen, meaning that approximately 45% of all District children use Instagram at least daily, and 76% of children use Instagram on a monthly basis. $^{10}$ Emily A. Vogels, et al., Teens, Social Media and Technology 2022, PEW RSCH. CTR. (Aug. 10, 2022), [https://www.pewresearch.org/internet/2022/08/10/teens-social-media-and-technology-2022](https://www.pewresearch.org/internet/2022/08/10/teens-social-media-and-technology-2022). $^{11}$ Id. $^{12}$ Id. * * * 53. In 2021, over 70% of all District high school students reported spending more than three hours of an average school day on screen time—in front of an electronic device or television, not counting time spent on schoolwork.$^{13}$ This figure is even more alarming when the statistic focuses on high school girls—74% of whom reported spending three or more hours on screens.$^{14}$ The numbers are similar for District middle school students—over two thirds (67.5%) of whom report spending three or more hours on screens.$^{15}$ To put that into perspective, at least two thirds of middle and high school-aged children in the District are spending the equivalent of half of their school day on screens. 54. Between 2021 and 2023, the number of active District teen Instagram users has skyrocketed. In 2021, an average of 41,217 teens in the District were daily active users of Instagram and 49,209 were monthly active users. Relative to District usage rates among children for April 2023, this means that daily Instagram usage has increased by 35% and monthly Instagram usage has increased by 92%, just in the last two years. II. Meta knowingly traps children into compulsive and excessive use of its Platforms, even though it knows it is contributing to a mental health epidemic. 55. This section explains (1) the features of Meta’s Platforms that harm and ensnare children; (2) the science behind how these features harm children; and (3) that Meta is, and has been, well aware of the harm it causes to children. $^{13}$ 2021 District of Columbia Youth Risk Behavior Survey Results: District of Columbia (Including Charter Schools) High School Survey, OFF. OF THE STATE SUPERINTENDENT OF EDUC., [https://osse.dc.gov/sites/default/files/dc/sites/osse/page\_content/attachments/2021DCBH%20Summary%20Tables.pdf](https://osse.dc.gov/sites/default/files/dc/sites/osse/page_content/attachments/2021DCBH%20Summary%20Tables.pdf). $^{14}$ Id. $^{15}$ Id. * * * ## A. Meta knows its Social Media Platforms contain addictive features designed to consume the time and attention of teenagers. 56. Meta develops and implements features intended to hook children into spending as much time as possible on its Social Media Platforms in order to maximize profits. Some examples of these manipulative design features are Meta’s personalization algorithms, alerts, infinite scroll, autoplay, and short-form videos known as “Reels.” ## i. Personalization Algorithms 57. Meta employs personalization algorithms universally across its Social Media Platforms, including in the Instagram platform’s Main Feed (the scrolling presentation of content immediately visible upon opening the app) and Explore Feed (another scrolling presentation of algorithmically-curated content that can be optionally guided by a user’s text input in a search field). 58. Meta changed Instagram’s user feeds in 2016 to incorporate this personalization algorithm. Prior to 2016, Instagram user feeds were simply in reverse chronological order. 59. Meta’s goal is to engage the user with its personalization algorithms. To achieve that goal, Meta’s personalization algorithms serve children categories of content based on a sequencing method referred to by psychologists as “variable reinforcement schedules” or “variable reward schedules.” 60. These variable reward schedules work by periodically—but not always, and not in a predictable pattern—delivering types of content that trigger a release of dopamine, a neurotransmitter released by the brain in response to certain stimuli. Dopamine, commonly seen to be the “pleasure chemical,” is released in anticipation of a potential reward. However, dopamine neurons fire for only a relatively short period of time, and after dopamine is released, an “individual can become disheartened and disengaged.” * * * 61. By algorithmically serving content to children according to variable reward schedules, Meta manipulates dopamine releases in its child users, inducing them to engage repeatedly with its products—much like a gambler at a slot machine. The gambler pulls the lever repeatedly, receiving pleasure from anticipating that this pull might lead to a reward—or the next one, or the next one. 62. But while a slot machine delivers rewards randomly, Meta induces the same pleasure by intentionally creating a randomized schedule for “rewarding” the user. In this way, it is more insidious than a slot machine. A slot machine is truly random, indifferent to the gambler’s profile or previous conduct, but Meta’s personalization algorithms are optimized to withhold and provide rewards according to a schedule tailored to ensure that the user craves more content and continues using the Platforms. Each new post or notification stirs up pleasure in the user as they anticipate receiving a reward. 63. This is particularly effective and dangerous for adolescents. Because their brains are still developing, they lack both impulse control and the full range of executive functions. There is a reason children are not permitted to gamble at slot machines. 64. Because the algorithms used in Meta’s Platforms are consistently learning more about users as the users scroll and search for more related, dopamine-inducing content to display, the algorithms tend to push users into a “rabbit hole” of content. The algorithms fixate on a particular interest that a particular user has, and the interest shows up in all their recommendation feeds and Reels—becoming more and more extreme beyond the user’s initial expressed interest and isolating the user from other thoughts or opinions. * * * ## 65\. Meta knows about this phenomenon. As a Meta employee stated, People don’t just fall into rabbit-holes (and we stopped using this term.) They have some preference that our models amplify (hence the term ‘preference amplification’) - eg, follow accounts, like reels, etc. Then they tend to ‘drift’ towards what the rec\[ommendation\] sys\[tem\] shows them, which is further picked up by the model, which makes the problem even worse. 66. When a user is in one of these “preference amplification” holes, the absence of other, more moderate information makes it difficult for the user to ever escape the hole—and the problems that staying increasingly on-screen has caused them. In 2021, Meta employees internally acknowledged how Instagram’s ranking algorithm takes youth “into negative spirals & feedback loops that are hard to exit from.” 67. Meta purposefully designs its personalization algorithm to lure users into passively consuming more content in search of their next dopamine hit. Meta does so fully aware that as a consequence of this design the algorithm pushes users into ever more extreme content in order to keep them engaged longer. ## ii. Alerts 68. Instagram, by default, employs a range of alerts when the application is installed on a smartphone. Meta carefully designs the alerts, which it enables by default, to increase engagement by children by taking advantage of well-understood neurological and psychological phenomena, including using sounds and vibrations to trigger sudden dopamine releases. 69. These include vibrating, visual, and sound alerts to a user’s mobile phone (some of which remain until viewed), notifications that show up in-app, and email notifications. 70. These alerts allow Meta to disrupt its users at any time of day to lure them back onto its Social Media Platforms. * * * 71. By default, Meta notifies users when other users they follow take any of the following actions: - following the user, - “Going Live” (i.e., starting a live broadcast), - liking or commenting on the user’s posts, - mentioning the user in a comment or tagging the user in a post, and - sending the user a message. 72. These alerts are disruptive for all users but are especially harmful for children, who are particularly vulnerable to distraction and psychological manipulation. Meta knows that “smartphone notifications caused inattention and hyperactivity among teens, and they reduced productivity and well-being.” Researchers have documented how these notifications impact the brain similarly to stimulating drugs. 73. These notifications prompt users to develop a compulsion to re-open and re-engage with Meta’s Platforms repeatedly throughout the day and at night when prompted. 74. These notifications are extremely effective. A recent study showed teens checked their phones on average fifty-one times per day, with some teens checking their phone over 400 times a day. During that same period, the teens in the study received a median of 237 notifications on their smart phones per day, with some users receiving as many as 4,500 in a single day. On average 23% of those notifications arrived during school hours and 5% during sleeping hours on school nights. As the study concluded, smartphone app developers could do a better job of eliminating notifications during times of day that are more disruptive to young people. * * * 75. While users can technically disable notifications, the addictive elements of these Platforms are a substantial barrier to children taking the steps needed to disable them. By designing its default settings with notifications on, Meta has taken advantage of this reality, creating a significant barrier to turning off notifications and leading many users to spend more time on its Social Media Platforms than they otherwise would. ## iii. Infinite Scroll and Autoplay 76. “Infinite scroll” is characterized by 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 (without seeing the top portion of the next post in their feed). The “teasing” of yet-to-be-viewed content continues indefinitely—as the user scrolls down the feed, new content is automatically loaded and “teased.” 77. Meta first implemented a form of this endless scrolling on Facebook. Since 2011, Facebook delivers users to an endless, seemingly random-ordered feed of content on its home page or “News Feed.” Prior to 2011, the News Feed was a chronological list of everything a user’s friends had been up to. In 2016, Instagram similarly debuted its “infinite scroll” system of delivering content. 78. The “infinite scroll” format makes it difficult for children to disengage, because there is no natural end point for the display of new information. The Platforms do not stop displaying new information when a user has viewed all new posts from their peers. Instead, the Platforms display new social content and suggest relevant information that has yet to be viewed, provoking the children’s fear of missing out, or “FOMO.” 79. Meta does not allow users to turn off infinite scroll. So long as they choose to use Facebook or Instagram, they are stuck with it. * * * 80. Meta also deploys the “autoplay” feature to keep children on its Platforms. Much like “infinite scroll,” the “autoplay” feature of Instagram’s and Facebook’s “Stories” encourages children to continuously engage on the platform because, once one Story is complete, the user is automatically moved to the next Story which starts playing immediately. The default programing keeps the user watching unless the user takes affirmative action to disengage. 81. While Facebook allows users to turn off autoplay, the setting is “on” by default. Autoplay is also on by default on Instagram, and disabling the feature is difficult. 82. Meta is keenly aware that teens are particularly susceptible to the infinite scroll and autoplay features. In August 2021, Meta notified its staff that YouTube turned off autoplay for users under the age of eighteen. Internally, Meta staff noted: “Turning off autoplay for teens seems like a huge move! Imagine if we turned off infinite scroll for teens.” ## iv. Ephemeral Content 83. In 2016, Meta also started implementing “ephemeral content” features in its Social Media Platforms to further induce a sense of FOMO in children and keep them checking the Platforms. Ephemeral content is made only temporarily available to users with notifications and visual design cues indicating that the content will soon disappear forever—encouraging users to frequently check their social media accounts. 84. One example of ephemeral content is the “Stories” feature, which Meta added to Facebook and Instagram in 2016, specifically to drive up teen engagement. 85. Meta designed the “Stories” features to show content for only a short amount of time before disappearing from the feed. This causes children to frequently open Meta’s Social Media Platforms so they do not “miss out” on any new content. * * * 86. Another example of ephemeral content is Instagram Live, also introduced in 2016, where content creators can live-stream content that users can watch and react to in real time. Unlike other content delivery systems, which permit a user to view existing posts on a schedule convenient for the user, content released through “Live” is only available in real time, such that a young user’s failure to quickly join the live stream when it begins means that the user will miss out entirely on the chance to view the content. 87. Meta makes sure that its users are notified of the potential to miss new Live content, sending a push notification to interested users that reads, “\[@user\] started a live video. Watch it before it ends!” 88. An executive summary circulated internally in 2016 regarding Live content indicated the goal is to increase the time users—specifically, teens—spend watching these videos. 89. But Meta does not have to make these videos disappear. Rather, Meta chooses to use ephemeral content to induce a sense of urgency in children, so they feel a need to return to the Platform—or miss out forever. ## v. Reels 90. In 2020 and 2021, Meta upped the addictive design of its Social Media Platforms with the introduction of Reels. Reels uses Meta’s algorithms to present short-form videos individually tailored to users’ interests; based on data collected from each user, Meta predicts how and whether the user will engage with each Reel. Reels then spoon-feeds children an infinite stream of short videos perfectly suited to monopolize their shorter attention spans. 91. Like infinite scrolling, Reels automatically and perpetually play as the user swipes the screen up to the next video. The short-form nature of Reels (between fifteen to ninety seconds, as of April 2023), and the frameless way it fills a user’s screen, ensures that the user will not get bored and navigate away or close the app. 92. Meta deployed Reels to compete with competitors, like TikTok, which offer similar features and are growing in popularity. Meta’s investment in Reels was also specifically designed to attract and increase youth engagement. * * * 93. Meta designed and implemented all the foregoing features and algorithms while acutely aware that children are particularly susceptible to these types of manipulations and that an increase in compulsive, excessive, and harmful use would foreseeably result. But Meta implemented and continues to maintain those features and algorithms anyway. 94. As the inventor of the infinite scroll feature, Aza Raskin, retrospectively reasoned in 2018: It’s as if they’re taking behavioural cocaine and just sprinkling it all over your interface . . . \[b\]ehind every screen on your phone, there are generally like literally a thousand engineers that have worked on this thing to try to make it maximally addicting . . . In order to get the next round of funding, in order to get your stock price up, the amount of time that people spend on your app has to go up . . . \[s\]o, when you put that much pressure on that one number, you’re going to start trying to invent new ways of getting people to stay hooked. ## B. Compulsive and excessive use of Meta’s Social Media Platforms harms children. 95. Social media use among children, teens, and young adults dramatically increased in the United States beginning in 2012, when Meta acquired Instagram. Instagram went from 50 million users in 2012 to over 600 million users by 2016. 96. Increased use of social media platforms, including those operated by Meta, results in psychological and health harms among children, including increased rates of major depressive episodes, anxiety, sleep disturbances, suicide, and other mental health concerns. See Figure 3, showing the proportion of adolescents with depression or low psychological well-being, by hours a day of social media or smartphone use. (Figure 3.)16 97. An internal Meta presentation from March 2019 recognizes that 55% of Meta’s users in the U.S. suffered from problematic use, which Meta described as “serious, negative impacts on sleep, relationships, work, or lives, combined with lack of control over FB use,” and that 3.1% of its users suffered from “severe” problematic use. 98. These harms are particularly acute in children. 16 Jean M. Twenge, Increases in Depression, Self-Harm, and Suicide Among U.S. Adolescents After 2012 and Links to Technology Use: Possible Mechanisms, 2:1 PSYCHIATRIC RES. AND CLINICAL PRACTICE 19, 22 (Mar. 27, 2020), [https://prcp.psychiatryonline.org/doi/10.1176/appi.prcp.20190015](https://prcp.psychiatryonline.org/doi/10.1176/appi.prcp.20190015). * * * 99. During adolescence, risk-taking behavior is at its peak and self-esteem is at its most vulnerable. Brain regions associated with a desire for risk-taking, attention, peer feedback, and reinforcement become particularly sensitive in adolescence, while the regions associated with maturity and impulse control are not fully developed until adulthood. Because identities and sense of self are not yet fully formed, teens are more susceptible than anyone else to the danger, misinformation, peer pressure, and false images that abound on social media. 100. The brain goes through massive changes during adolescence, both maturing and even changing its actual structure. These changes are what help teenagers turn—eventually—into functioning adults who are self-motivated, emotionally mature, less impulsive, and able to manage their own needs (and eventually those of a household). But at the same time the brain is still developing its impulse control and emotional stability, activity in the area of the brain that feels pleasure in response to rewards is at its peak. This is the exact area that reward patterns like Meta’s stimulate. When stimulated, this area of the brain releases dopamine, emitting a feeling of pleasure. The mismatch in brain maturation means that adolescents are highly susceptible to risky behaviors and temptations that can prove damaging to their development and well-being. 101. As a U.S. Surgeon General advisory recognized, “the current body of evidence indicates that while social media may have benefits for some children and adolescents, there are ample indicators that social media can also have a profound risk of harm to the mental health and well-being of children and adolescents.” 102. Moreover, Meta can maintain any benefits the Platforms offer without the manipulative features they utilize to drive up engagement. On multiple occasions, Meta has 17 U.S. DEP’T OF HEALTH & HUM. SVCS., Social Media and Youth Mental Health: The U.S. Surgeon General’s Advisory 4 (2023), [https://www.hhs.gov/sites/default/files/sg-youth-mental-health-social-media-advisory.pdf](https://www.hhs.gov/sites/default/files/sg-youth-mental-health-social-media-advisory.pdf). * * * explicitly considered and explicitly rejected design changes after finding that those changes would decrease the harms its Platforms inflict, but also decrease engagement—and therefore its profits. 103. Children who use social media for at least five hours per day are many times more likely to have clinically relevant symptoms of depression than non-users. Heavy users of social media, at four or more hours a day, may emerge from puberty stunted or otherwise damaged, perhaps permanently. Indeed, frequently checking social media has been associated with distinct changes in the developing brain in the amygdala, which is vitally important for impulse control and emotional regulation, and could increase adolescent sensitivity to reward and punishment. 104. Making matters worse, heavier social media use was associated with poor sleep patterns in adolescents (e.g., later sleep and wake times on school days and trouble falling back asleep after nighttime awakening) and poorer sleep quality. Poor sleep in turn causes or exacerbates symptoms of depression and anxiety. Lack of sleep also has negative physical effects, slowing down bodily functions like immune response and metabolism. 105. Young people can be particularly attuned to developing a fear of missing events or experiences (i.e., FOMO) when they are not online and may feel an extra need to be connected at night and to check social media. Many teens frequently wake up at night to check social media notifications. 106. Meta has been aware since at least 2019 that excessive social media use is correlated with sleep problems, which negatively impacts mental health. Yet Meta still continues to use notifications and other practices that disrupt sleep. 107. A recent study showed that teens received a median of 237 notifications on their smart phones per day, with some users receiving as many as 4,500 in a single day. On average * * * 23% of those notifications arrived during school hours and 5% during sleeping hours on school nights. As the study concluded, app developers, such as Meta, could do a better job of eliminating notifications during times of day that are more disruptive to young people. 108. In 2021, nearly three-quarters (73.6%) of District teens reported getting fewer than eight hours of sleep on an average school night. During that same period, over 70% of District high schoolers reported three hours or more of screen time on an average school day.$^{18}$ 109. Habitual social media use also affects how children’s brains mature, and habitual social media users’ brains develop differently than non-habitual users in many key areas. 110. Researchers have identified a feedback loop: those who use social media habitually are less able to regulate their behavior. That habitual use and decreased ability to self-regulate, in turn, can lead back to more social media use. And, restarting the cycle, that additional use makes it even harder to regulate the problematic behavior. 111. The harms associated with habitual or prolonged use are by no means hypothetical. Young people in the U.S. are in a mental health crisis. During the same period social media use increased, young people began suffering severe mental health harms in greater numbers. Data from the Centers for Disease Control and Prevention shows that the percentage of high school students “who experienced persistent feelings of sadness or hopelessness” skyrocketed between 2013 and 2021.$^{19}$ $^{18}$ 2021 District of Columbia Youth Risk Behavior Survey Results: District of Columbia (Including Charter Schools) High School Survey 73-80, OFF. OF THE STATE SUPERINTENDENT OF EDUC., [https://osse.dc.gov/sites/default/files/dc/sites/osse/page\_content/attachments/2021DCBH%20Summary%20Tables.pdf](https://osse.dc.gov/sites/default/files/dc/sites/osse/page_content/attachments/2021DCBH%20Summary%20Tables.pdf). $^{19}$ Youth Risk Behavior Survey, Data Summary & Trends Report: 2011-2021 62, CTRS. FOR DISEASE CONTROL AND PREVENTION (2023), [https://www.cdc.gov/healthyyouth/data/yrbs/pdf/YRBS\_Data-Summary-Trends\_Report2023\_508.pdf](https://www.cdc.gov/healthyyouth/data/yrbs/pdf/YRBS_Data-Summary-Trends_Report2023_508.pdf). * * * 112. Risk of suicide is up, too. In 2011, nineteen percent (19%) of high school girls seriously considered attempting suicide. By 2021, that figure reached thirty percent (30%). Adolescents aged twelve to seventeen and girls saw the greatest increases in suicidal ideation and attempts in the same time period.$^{20}$ Indeed, in 2013 alone, the suicide rate for thirteen-year-old girls jumped by fifty percent (50%).$^{21}$ 113. The mental health crisis among youth nationally is also affecting the District, which has witnessed an alarming worsening of mental health trends among adolescents since 2007. 114. In 2021, almost half (47.7%) of District high school girls self-reported persistently feeling sad or hopeless.$^{22}$ The rates are similarly high for the District’s high school students overall, with more than a third (36.3%) self-reporting feelings of sadness or hopelessness.$^{23}$ Since 2007, the number of District high school students reporting feeling sad or hopeless has increased by a total of 28.3% overall and by 47.2% among high school girls. Similarly, self-reports of suicidal thoughts among youth have also increased in the District. In 2021, over a quarter (28%) of all middle school students reported seriously considering attempting suicide. Those numbers are even worse for middle school girls, over a third (36.9%) $^{20}$ Gregory Plemmons, et al., _Hospitalization for Suicide Ideation or Attempt: 2008-2015_, 141:6 PEDIATRICS 141 (June 1, 2018), [https://doi.org/10.1542/peds.2017-2426](https://doi.org/10.1542/peds.2017-2426); see also Brett Burstein, et al., _Suicidal Attempts and Ideation Among Children and Adolescents in US Emergency Departments, 2007-2015_, 173 JAMA PEDIATRICS 598, 598-600 (2019), 10.1001/jamapediatrics.2019.0464. $^{21}$ Jonathan Haidt, et al., _Social media and mental health: A collaborative review_, NEW YORK UNIV. (unpublished manuscript), [https://docs.google.com/document/d/1w-HOfseF2wF9YIpXwUUtP65-olnkPyWcgF5BiAtBEy0/edit](https://docs.google.com/document/d/1w-HOfseF2wF9YIpXwUUtP65-olnkPyWcgF5BiAtBEy0/edit) (last visited Oct. 17, 2023). $^{22}$ 2021 District of Columbia Youth Risk Behavior Survey Results: District of Columbia (Including Charter Schools) High School Survey, OFF. OF THE STATE SUPERINTENDENT OF EDUC., [https://osse.dc.gov/sites/default/files/dc/sites/osse/page\_content/attachments/2021DCBH%20Summary%20Tables.pdf](https://osse.dc.gov/sites/default/files/dc/sites/osse/page_content/attachments/2021DCBH%20Summary%20Tables.pdf). $^{23}$ Id. * * * of whom reported seriously considering attempting suicide in 2021. This is nearly a 25% increase in suicidal ideation among middle school girls since 2007.$^{24}$ The increase for high school girls is even worse. In 2007, 16.3% of high school girls reported seriously considering suicide, while in 2021 that number was over a quarter (25.2%), representing a 54.6% increase. (Figure 4. Aggregated data from the District of Columbia’s Youth Risk Behavior Survey Results between 2007-2021.) (Figure 5. Aggregated data from the District of Columbia’s Youth Risk Behavior Survey Results between 2007-2021.) $^{24}$ Id. * * * 115. Notwithstanding Meta’s knowledge of the harms excessive and compulsive use of its Social Media Platforms causes children, it continues to implement addictive features to induce prolonged use for its own profit. ## C. Meta knows that children use its Social Media Platforms excessively and compulsively to the detriment of their mental health, yet it refuses to address the problem. 116. Meta exploits the psychological vulnerabilities of children through the false promise that meaningful social connection lies in the next Story, Reel, image, or video—and that ignoring the next piece of social content could lead to social isolation. Meta is aware that teens are more susceptible to compulsive and excessive use of its Social Media Platforms. Meta is also aware that extreme users of Instagram (more than five hours per day) are more likely to be teens. 117. In 2020, internal Meta statistics showed that teens are 40% more likely to spend five or more hours on Instagram per day than non-teens. Over the course of a week, teens are also 40% more likely to spend more than twenty-eight hours on Instagram than non-teens. 118. It was also widespread knowledge within Meta that teens were particularly susceptible to the addictive designs Meta employed to keep them engaged. Over the years, Meta employees have delivered many presentations, conducted many studies, reviewed extensive literature, and written many articles on the sociopsychological vulnerabilities unique to teenagers that keep them engaged on Meta’s Platforms. 119. Meta is well aware that “\[t\]een brains are much more sensitive to dopamine” compared to adult brains. Meta’s internal research has concluded that: “Teen brains are much more sensitive to dopamine, one of the reasons that the risk of drug addiction is higher for adolescents and it’s the same thing that keeps them scrolling and scrolling. Due to the immature brain, they have a much harder time stopping even though they want to—our own product foundation research has shown teens are unhappy with the amount of time they spend on our app.” 120. In 2020, Meta started a project focused on deepening the company’s understanding of teens, including brain development. In a lengthy presentation, the company outlined “over 80 internal studies” and external research on adolescent brain development, concluding that because teen frontal cortexes are not fully developed, “teens are highly dependent on their temporal lobe where emotions, memory and learning, and reward system reign supreme.” 121. The company also recognized that “\[t\]een decisions and behavior are mainly driven by emotion, the intrigue of novelty and reward . . . they make teens very vulnerable . . . \[e\]specially in the absence of a mature frontal cortex to help impose limits on the indulgence in these.” 122. Meta designs its Platforms to increase engagement by manipulating children’s dopamine levels. As Meta itself stated: “Teens are insatiable when it comes to ‘feel good’ dopamine effects,” and “Teen’s insatiable appetite for novelty puts them on a persistent quest to discover new means of stimulation.” 123. Meta researchers know that children are using the platform to their detriment, with compulsive and excessive use interfering with daily activities, including sleep. 124. Meta researchers noted that teens “talk about the amount of time they spend on Instagram as one of the ‘worst’ aspects of their relationship to the app.” Meta researchers observed that in conversations, teens had “an addicts’ narrative about their use” and “wish\[ed\] they could spend less time caring about it, but they can’t help themselves.” * * * 125. In 2020, a Meta employee noted that “\[t\]he feedback, essentially, is that (1) teens feel addicted to IG \[Instagram\] and feel a pressure to be present, (2) like addicts, they feel that they are unable to stop themselves from being on IG, and (3) the tools we currently have aren’t effective at limiting their time on the ap\[p\].” 126. Meta has described Instagram as “deliver\[ing\] \[teens\] a dopamine hit” every time a teen “finds something unexpected” on the app, fulfilling their brains’ “insatiable” need for “feel good dopamine effects,” to which “teen brains are much more sensitive.” 127. Based on multiple internal studies, Meta has known for years that that the way Meta designs its Social Media Platforms (and not merely the content on the Platforms) causes children to compulsively use its Platforms, thereby contributing to children’s poor mental health. 128. Meta’s own research confirms that its Social Media Platforms are among the worst in harming children. A 2021 study showed that 6% of teen girls in the U.S. and 13% of teen girls in the UK “traced their desire to self-harm/commit suicide to Instagram.” 129. Meta knows that the compulsive and excessive social media use that it actively promotes and profits from is detrimental to children. Meta also knows that children are facing a mental health crisis. According to Meta’s own studies, “82% of teens have felt at least one emotional issue in the past month. One in five has thought about suicide or self-injury.” 130. Meta itself acknowledges that Instagram provides no value to teens in return for their time: “\[t\]he endless content is addicting, wasting time that could be spent doing something more productive.” 131. Meta is aware that the majority of its manipulative design features encourage “passive” use of its Social Media Platforms, viewing other people’s content rather than posting and generating their own content. Meta employees admit that particular features like Reels and * * * Stories encourage passive consumption, and that passive consumption is generally worse for user well-being. 132. Many Meta researchers have established that passive consumption of social media content harms mental health. Meta also knows that children dislike passive consumption sessions on Meta’s Platforms, but still feel compelled to engage in them. 133. Since 2018 Meta has touted so-called “time-management tools” on its Social Media Platforms as a way to help users control their experience on the Platforms, even knowing they were ineffective to counteract the problem of compulsive and excessive use it had created and continued to fuel. Meta was aware that its time-management tools were only being used by a tiny percentage of users. Rather than prevent excessive use, these measures provided a notification to users giving them the option to stop scrolling, which most users dismissed. 134. Meta routinely chooses to prioritize profit over the safety and well-being of its users. Instead of the ineffective time-management tools it touted, Meta could have—but refused to—take steps that its own research shows could have actually limited its Platforms’ harm to children. For example, Meta could have removed the addictive design features and algorithms it implemented. Meta has staunchly refused to take such action for the sake of profit, even when presented with various proposals to mitigate its Platforms’ harms to children. But time and again, Meta either failed to act or implemented insufficient half-measures—in multiple instances at the direction of senior leaders. * * * ## III. Meta knowingly misrepresents that its Social Media Platforms are safe and knowingly conceals that they are not. 135. At the same time Meta continues to develop and implement features that induce children’s extensive, compulsive, and harmful social media use, it constantly and deceptively reassures parents, lawmakers, and users that its Social Media Platforms are suitable for children and designed to promote their well-being. 136. For years, Meta has publicly claimed that its top priority is well-being and that its Platforms are safe and age-appropriate platforms for children. However, Meta has known these claims are misleading and continually chooses to maximize profits without limits over the health and safety of children when making decisions regarding the design and management of its Social Media Platforms. 137. Ultimately, Meta’s leaders (including Mark Zuckerberg) have final say over its Social Media Platforms, and those leaders have refused to make changes to reduce these harms where changes might also affect Meta’s profits. 138. When faced with a choice between user engagement and user well-being, Meta consistently chooses the former, while representing to consumers it is choosing the latter. ## A. Meta knowingly misrepresents that its Social Media Platforms are not designed to hook children. 139. Although one of Meta’s key goals is to induce children to spend ever-increasing amounts of time on its Social Media Platforms, the company has vehemently denied this. Instead, Meta represents its Platforms are safe and age-appropriate for children, despite using tactics known to harm them. 140. Meta’s 2023 Responsible Business Report represented that it actively protects mental health on its Platforms: “At Meta, protecting and supporting digital well-being is always a top priority. We want people to connect with others in a safe, positive and supportive environment and leave our apps feeling good about the time they spend on them.”25 141. In 2018, when asked whether Meta studied dopamine feedback loops to keep users trapped on its Platforms, Mark Zuckerberg responded: “No . . . that’s not how we talk about this or how we set up our product teams. We want our products to be valuable to people, and if they’re valuable, then people choose to use them.”26 142. In 2018, then-COO Sheryl Sandberg’s prepared talking points to reporters included the claim that Meta “do\[es\] not optimize \[its\] systems to increase amount of time spent in News Feed” and “explicitly do\[es\]n’t give \[its\] teams goals around time spent.” 143. In October 2019, Zuckerberg publicly stated that Meta does not allow Meta “teams \[to\] set goals around increasing time spent on \[Meta’s\] services.” 144. Additionally, Zuckerberg’s talking points for his 2021 testimony to Congress included the misleading statements: “The impact that our products have on the wellbeing of everyone is a top priority. It’s not how much time you spend online, it’s how you spend it.” The talking points also falsely deny that Meta designs its products “to be addictive so people spend more time on it and view more ads.” 25 2023 Responsible Business Practices Report 77, META (July 12, 2023), [https://scontent-ord5-](https://scontent-ord5-/) 2.xx.fbcdn.net/v/t39.8562- 6/10000000\_1974692352892843\_2880908794103790486\_n.pdf?\_nc\_cat=107&ccb=1- 7&\_nc\_sid=e280be&\_nc\_ohc=MtAD9xN692sAX\_pgWbG&\_nc\_ht=scontent-ord5- 2.xx&oh=00\_AfDwYsMrH0EH-qc9bK-dbi2Ag8jjVvMebykTG-SfwOkbRg&oe=653A6845. 26 Facebook, Social Media Privacy, and the Use and Abuse of Data: Hearing Before the S. Comm. on Commerce, Sci., and Transp. and H. Comm’s on the Judiciary and Commerce, Sci., and Transp., 115th Cong. (Apr. 10, 2018), available at [https://www.commerce.senate.gov/2018/4/facebook-social-media-privacy-and-the-use-andabuse-of-data](https://www.commerce.senate.gov/2018/4/facebook-social-media-privacy-and-the-use-andabuse-of-data). * * * 145. The reality is that Meta closely tracks several data points related to its teen users, including daily average use and number of sessions for daily users. Meta’s goal is to keep its userbase growing and increase the time teen users spend on its Platforms. In evaluating design features such as infinite scroll and autoplay, Meta focuses on whether they increase the time users spend on the Platforms. 146. Meta’s focus on increasing time spent by teens goes back many years. It has consistently judged the success of its Platforms by retention and increase of “teen time spent” since as early as 2016. 147. As discussed in detail above, Meta has spent significant time studying the neurological development of adolescents, conducting over eighty internal studies on the subject. Meta has used this extensive internal research “to inform product strategy” and increase teen engagement. 148. Sean Parker, founding president of Meta, explicitly acknowledged that the purpose of Meta’s Social Media Platforms is to consume children’s time: The thought process that went into building these applications, Facebook being the first of them . . . was all about: “How do we consume as much of your time and conscious attention as possible?” That means that we need to sort of give you a little dopamine hit every once in a while, because someone liked or commented on a photo or a post or whatever. And that’s going to get you to contribute more content and that’s going to get you … more likes and comments . . . It’s a social-validation feedback loop . . . exactly the kind of thing that a hacker like myself would come up with, because you’re exploiting a vulnerability in human psychology. The inventors, creators—me, Mark \[Zuckerberg\], Kevin Systrom on Instagram, all of these people—understood this consciously. And we did it anyway. * * * ## B. Meta knowingly misrepresents the appropriateness of its Social Media Platforms for children and knowingly conceals the risks of exposing children to harmful content. 149. Despite its public representations about prioritizing user safety and shielding children from inappropriate content, Meta knows that it serves to children content that is not appropriate for their age. Yet Meta does not disclose this information to users or their parents. Externally, Meta leadership claims that it “age-gates” inappropriate content for children. “Age-gating” refers to the act of controlling or limiting access based on a user’s age. 150. In 2021, Meta’s Global Head of Safety Antigone Davis went as far as to say, “we don’t allow young people to see certain types of content. And we have age gating around certain types of content.” Davis also testified: “When it comes to those between 13 and 17, \[w\]e consult with experts to ensure that our policies properly account for their presence, for example, by age-gating content.” 151. That same year, Meta’s internal documents reveal that for children, only one out of every fifty pieces of content is actually “age-appropriate nutritious, or the sort of content we would like to promote to teens. This should also suggest that we have a big hill to climb on ensuring a higher % of a teen content consumption diet is ‘nutritious’…\[and\] also with developing new policy, measurement, detection, parental controls, and enforcement.” 152. Meta’s public communications about whether its personalization algorithms suggest and amplify distressing and problematic content are also false and misleading. Meta has vehemently denied that its Social Media Platforms amplify extreme, distressing, and/or problematic content. 153. However, Meta, in fact, does the opposite. Meta increases children’s engagement with its Platforms by periodically presenting them with psychologically and emotionally gripping content, including content related to eating disorders, violent content, content encouraging negative self-perception and body image issues, bullying content, and other categories of content known by Meta to inspire intense reactions from users. 154. In 2018, Meta internally recognized that the more time a user spends on Instagram, the more likely they are to be exposed to admissions of suicide and self-harm. In 2021, Meta employees internally acknowledged how Instagram’s ranking algorithm takes tweens “into negative spirals & feedback loops that are hard to exit from.” 155. That same year, internal Meta research showed that when a user followed anorexia-related content the personalization algorithms would recommend additional anorexiarelated content. 156. Meta knows that 13.5% of teen girls on Instagram say the platform makes thoughts of suicide and self-harm worse, 17% of teen girls on Instagram say the platform makes eating issues worse, and one in three teen girls say Meta makes their body image issues worse. 157. Meta understands the psychologically manipulative nature of its Platforms’ functionality, has knowledge that its personalization algorithms promote harmful content, and is aware that users “wish\[\] Instagram \[gave\] them better control over what \[content\] they \[see\].” 158. Even though Meta knows that its products are harmful to teenagers’ mental health, Meta externally characterizes Instagram as a source of support for teens struggling with thoughts of suicide and self-harm, and mental health issues generally. 159. Meta’s deception regarding its personalization algorithms’ promotion and amplification of harmful content deprives users (and the parents of children who use the Platforms) of informed decision-making regarding whether (and how) to engage with Meta’s products and Social Media Platforms. * * * ## C. Meta knowingly conceals the existence and prevalence of adult predators who contact children via its Platforms. 160. Meta has been aware for years that adult-predatory contact is an issue on its Social Media Platforms and it has failed to effectively counteract the problem. Because it fails to disclose these known risks to children and their parents, Meta allows them a false sense of security. 161. Meta is well aware that predatory adults pretend to be teens on Meta’s Platforms to interact with minors and prepare them for “sextortion or CEI \[child exploitive imagery\] trading.” 162. In 2020, Meta released an internal report cataloging the Platforms’ “Child Safety Issues,” expressly finding that: (1) Instagram had “minimal child safety protections”; (2) the company’s “\[m\]inor sexualization policies and classifiers \[are\] immature”; (3) interoperability of the Platforms “increases discoverability and reachability of minors across apps, need stronger protections particularly for IG \[Instagram\]”; (4) “\[l\]ive video chat and remote presence products introduce new vector for abuse, lack reporting with content capture”; and (5) ephemeral content “impacts efficacy of reporting.” 163. Following the report, internal recommendations were circulated noting that “work is in progress to block unconnected adults on \[Facebook\] and \[Instagram\] from initiating messages with minors on \[Facebook\], however work is not planned to prevent \[Facebook\] and \[Instagram\] adults from initiating messages to \[Instagram\] minors.” No reason was provided for this distinction; additionally, it does not appear that Meta ever implemented the “block” discussed for Facebook minors. Nor did Meta look to implement other effective methods of reducing this contact, because “product leadership has been unwilling to prioritize these mitigations,” or the company allegedly lacked resources and strong leadership buy-in to “accomplish even bare minimum protections.” 164. In May 2021, Meta received external research completed on social media platforms including Instagram and Facebook. This research revealed that children as young as nine years old were using Meta’s Platforms and having online sexual interactions with adults. 165. Meta knows that Instagram’s direct messenger (“DM”) has thirty-eight times the amount of unwanted illicit contact between minors and adults as Facebook, yet enforcement and takedowns on Instagram are drastically lower. More generally, Meta knows that “a top privacy problem (based on prevalence and severity) on Instagram is receiving messages from unwanted people.” 166. Despite this knowledge, Meta does not even take the most basic step of telling children or their parents about the risks they face by simply using Instagram. D. Meta knowingly misleads consumers, including parents, about features it knows promote body dysmorphia and eating disorders in children. 167. Meta also deceives consumers by representing that its Social Media Platforms do not allow content that promotes or encourages eating disorders—all while actively choosing to retain platform features known by Meta to promote those very problems. 168. Meta tells parents that it takes steps to combat body dissatisfaction and eating disorder content on its Platforms. Meta claimed in its “Parent’s Guide” that it published on its website for “parents with teens who use Instagram” that it “work\[s\] with experts to help inform our product and policies” around eating disorders.$^{27}$ $^{27}$ Marne Levine, _A New Resource for the Parents of Teens on Instagram_, INSTAGRAM (Sep. 6, 2018), [https://about.instagram.com/blog/tips-and-tricks/new-resource-for-what-parents-need-to-know-about-instagram](https://about.instagram.com/blog/tips-and-tricks/new-resource-for-what-parents-need-to-know-about-instagram); _A parent and guardian’s guide to Instagram_ 63, INSTAGRAM, * * * 169. Similarly, in 2021, Meta’s Global Head of Safety Antigone Davis testified before Congress and denied that Meta promotes harmful content to youth, such as content promoting eating disorders: “we do not direct people towards content that promotes eating disorders. That actually violates our policies, and we remove that content when we become aware of it. We actually use AI to find content like that and remove it.” 170. However, Meta deploys and makes available visual selfie camera filters that simulate facial plastic surgery available on its Social Media Platforms that it knows are “actively encouraging young girls into body dysmorphia.” 171. After public backlash in 2019, Meta’s initial response was to institute a temporary ban on the camera filters. Experts and Meta’s own employees recommended that the ban be made permanent or that Meta at least limit these filters to users over eighteen. 172. The ban was precipitated by concerns that these filters could have serious impacts related to body dysmorphia and eating disorders, especially for teenage girls. 173. Meta consulted over 20 outside experts and academics, and was told that “\[c\]hildren are particularly vulnerable” to these impacts, as well as “those with a history of mental health challenges \[and\] eating disorders\[.\]” 174. Prior to the ban, teenage girls were known to heavily use these cosmetic surgery filters. Meta’s own internal research found that “\[a\]mong teen girls that said they felt bad about their bodies, 32% said Instagram made them feel worse.” [https://scontent-ord5-2.xx.fbcdn.net/v/t39.8562-6/221635064\_179316607565042\_7019651825457490520\_n.pdf?\_nc\_cat=105&ccb=1-7&\_nc\_sid=e280be&\_nc\_ohc=isnomQtxpCAAX9sKwnH&\_nc\_ht=scontent-ord5-2.xx&oh=00\_AfDmfCQ52PFsKr4iV08JiiyDbBT5ZM2OqprK3N6oVU5U4Q&oe=6539D768](https://scontent-ord5-2.xx.fbcdn.net/v/t39.8562-6/221635064_179316607565042_7019651825457490520_n.pdf?_nc_cat=105&ccb=1-7&_nc_sid=e280be&_nc_ohc=isnomQtxpCAAX9sKwnH&_nc_ht=scontent-ord5-2.xx&oh=00_AfDmfCQ52PFsKr4iV08JiiyDbBT5ZM2OqprK3N6oVU5U4Q&oe=6539D768) (last visited Oct. 23, 2023). * * * 175. Despite these concerns, Mark Zuckerberg unilaterally vetoed the proposal to permanently ban plastic surgery simulation camera filters, stressing that there was a “clear\[\] demand” for the filters, and wrongly asserting that he had seen “no data” suggesting that the filters were harmful. In his view, “\[i\]t has always felt paternalistic . . . that \[Meta has\] limited people’s ability to present themselves in these ways.” 176. Later that week, a senior Meta staffer memorialized her disagreement with Zuckerberg’s decision, stating “I respect your call on this and I’ll support it, but want to just say for the record that I don’t think it’s the right call given the risks . . . I just hope that years from now we will look back and feel good about the decision we made here . . . I wanted to take a moment to express my disagreement even as we move to implement this decision. Thanks for listening.” 177. As of October 2023, these filters remain available on Instagram, and Meta has continued to promote these types of beauty-enhancing filters across its Platforms. Meta has even considered making “enhancement filters,” such as a “skin-smoothing” effect, a so-called “sticky setting” which would be automatically applied when users opened the camera after having used the filter once. 178. Having chosen to retain these harmful features, Meta has, at a minimum, an obligation to disclose these risks and harms to the public. ## E. Meta knowingly misleads consumers about the harms children face from its Social Media Platform features that encourage social comparison. 179. Meta has long said that it keeps its Platforms safe and appropriate for children to use—but it knows that this is not true. When given the chance to make the Platforms better for children, it seeks ever-greater profit instead. 180. Internally, Meta has said that any changes to improve teen safety and well-being must be balanced with teen engagement (i.e., the related effect on Meta’s revenue source). * * * 181. Meta’s Social Media Platforms contain design features that exacerbate teenagers’ comparison of themselves to others based on popularity or appearance, such as “likes,” a quick way for users to select other users’ photos and express validation or approval by clicking or tapping a heart icon, or the iconic thumbs-up icon. “Likes” were developed between 2010 and 2013. 182. Meta knows that adolescents using Meta’s Platforms tend to compare themselves to peers, and that “social comparison” and pressure to be perfect on Instagram have significant negative impacts, particularly on teenagers’ mental health. For appearance-based comparison, where a teen focuses on the physical looks of another user, teen girls aged thirteen to eighteen face the highest rates of negative comparison—that is, these teen girls are set up to find themselves less attractive or desirable than the other user. 183. Meta is also aware that its own design features are exacerbating this problem. Meta has internally recognized that teens suffered from constant negative comparisons on Instagram because Meta recommended triggering content to them and continued showing them “like” counts. 184. Meta’s senior leadership also admits that social comparison is a critical issue for its users, particularly on Instagram. Adam Mosseri (Head of Instagram since 2018) wrote in an internal email, “I see social comparison as the existential question Instagram faces within the broader question of whether or not social media is good or bad for people.” Because of Instagram’s “focus on young people and visual communication,” its emphasis on beauty and fashion content, and a “marketing look and feel often biasing too polished,” Mosseri reasoned that “social comparison is to Instagram \[what\] election interference is to Facebook.” * * * 185. Meta’s leadership, including Mark Zuckerberg, have stopped the Platform from making changes that might reduce harm to children. The reasoning, of course, is that those features might affect Meta’s profits. Meta’s public statements create the deceptive impression that, when faced with a choice between features that promote addictive user engagement and features that promote user well-being, Meta would choose the latter, but in reality, Meta consistently chooses the former. 186. One key example was Meta’s choice to not implement Project Daisy. 187. In 2020, Meta ran a test program called “Project Daisy,” where the “like” counts on Instagram posts were hidden. There were 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). 188. While the pilot project was underway, Meta publicly touted the program as an example of the company’s priority of user well-being. As early as 2019, Mark Zuckerberg publicly announced that Meta was “testing removing like counts on Instagram and Facebook. We do this because we know that if we help people have meaningful interactions, they’ll find our services more valuable.” Throughout 2019 and 2020, Instagram publicly promoted Project Daisy as Instagram’s move to address the “potentially corrosive impact of social media” and using it as an example of how Instagram is “\[w\]illing to make changes that will reduce the amount of time people spend on Instagram if it makes them safer.” * * * 189. Both Daisy programs successfully “reduced the negative impact of seeing posts with many Likes.” An internal Meta email from August 2020 noted that Daisy resulted in “less social comparison” and that “negative social comparison decreases more over time.” Another Meta internal report in 2020 concluded that Project Daisy “had a statistically significant impact on reducing the frequency of ‘like’ comparison for teens and non-teens.” 190. Despite advice from experts as well as evidence that a platform without public quantification of “likes” would decrease negative social comparison, Meta did not remove the “likes” feature for children on its Platforms. Instead, Meta required each individual user to opt in if they wished to hide “like” counts from their feed. 191. Top leadership at Meta ultimately decided not to implement Project Daisy as a default setting, despite knowing that the opt-in setting made the change completely ineffective—less than 1% of users found the setting and turned it on. Meta decided not to implement Daisy as a default setting because it could potentially have a negative effect on engagement, and could have a 1% negative impact on Meta’s revenue. 28 Amy Chozick, This is the Guy Who’s Taking Away the Likes, N.Y. TIMES (Jan. 21, 2020), [https://www.nytimes.com/2020/01/17/business/instagram-likes.html](https://www.nytimes.com/2020/01/17/business/instagram-likes.html). * * * 192. When releasing the new “opt-in” version of Daisy, Instagram downplayed its failure to launch Daisy as promised, instead touting the new version of Daisy as a way Meta was giving users “more control on Instagram and Facebook.” Meta told consumers: “What we heard from people and experts was that not seeing like counts was beneficial for some, and annoying to others, particularly because people use like counts to get a sense for what’s trending or popular, so we’re giving you the choice.” 193. Rather than own its decision to not implement Daisy, Meta chose to conceal internal findings that Project Daisy had statistically significant impacts on negative social comparison for users. It pretended that its decision to not implement Daisy was based on user choice and not adverse impact on its revenue. 194. Internally, Meta expressly acknowledged that any changes to improve teen safety and well-being (such as Project Daisy) would need to be balanced with teen engagement (i.e., the related effect on Meta’s revenue source). 195. Even though Meta knew that the opt-in version of Project Daisy was ineffective, Meta leadership publicly touted it as an example of how Meta was making Instagram safer. 196. Meta also made misleading statements regarding why Daisy wasn’t implemented, stating falsely that Daisy was not as effective as Meta hoped it would be, and hiding that the true reason Meta abandoned Daisy was that it had potential negative impacts on engagement and revenue. 197. Meta’s own researchers recognized that Meta was playing fast and loose with its description of Daisy’s effects and the researchers raised concerns about Meta externally portraying that Daisy did not improve well-being, saying: “statements like ‘Daisy didn’t improve well-being’ are misleading.” * * * 198. Meta ignored its researchers and chose to misrepresent to consumers that it was making changes to the platform to make it safer for children when it knew those measures were ineffective. ## F. Meta knowingly misleads consumers regarding the safety of its Social Media Platforms for children by hiding, misrepresenting, and obscuring its internal research findings. 199. Meta has also taken steps to hide its internal knowledge of the harms its users experience on its Platforms. As regulators, researchers, and the public at large began to scrutinize the harmful effects of social media more closely, Meta adopted a false veneer of transparency while concealing its own knowledge about its Platforms’ harms. 200. Meta carefully shields the inner workings of its Platforms—and their harms—from the public eye. Meta has a history of denying access to researchers to stymie research that might cast its Social Media Platforms in a negative light. 201. As part of the effort to conceal its internal findings, Meta has misled external researchers, regulators, and the public about the existence of research documenting the Social Media Platforms’ harms. In meetings with researchers regarding the effects of social media on teens’ mental health, Meta purposefully withheld internal research showing adverse findings. 202. In September 2021, Meta employees criticized the company for emphasizing that research into social media’s impact is inconclusive, when in fact Meta had conducted research on this issue with much more definitive findings. An employee even likened Meta’s public statements to those of tobacco companies, which similarly (and for years) relied on uncertainty in scientific studies to deny that cigarettes caused cancer. 203. Meta has internally conducted research showing that its Social Media Platforms are likely harming a significant portion of its userbase. Examples of these studies include: * * * 204. Compulsive Use: An internal study conducted in fall 2019 concluded that “Instagram is addictive, and time spent on \[the\] platform is having a negative impact on mental health.” The research observed that teens “have an addicts’ narrative about Instagram use . . . they wish they could spend less time caring about it, but they can’t help themselves.” The research concluded, “\[o\]n an everyday level, Instagram is a recipe for low level mental anxiety that unchecked can ladder up to something more serious.” 205. Mental Health: In the fall of 2019, Meta internally distributed results from a 22,000-person survey of Instagram users that found “at least 1 in 2 \[Instagram\] users had experienced at least one mental health related issue in the last 30 days.” 206. Body Dissatisfaction: In 2020, Meta internally distributed a Meta-sponsored literature review which found that “\[s\]ubstantial evidence suggests that experiences on Instagram or Facebook make body dissatisfaction worse,” and that users “perceived body image as a problem that Instagram worsened the most, more than when they end a relationship or lose a job.” 207. Negative Social Comparison: A spring 2020 internal survey of 100,000 individuals concluded that “\[a\]bout 1 out of 10 people experience negative social comparison on Instagram often or always,” and “1 in 4 people think that Instagram makes social comparison worse.” The study also concluded that this negative comparison is more common among teens, particularly teen girls. 208. Negative Social Comparison: In the fall of 2021 a report was internally distributed that concluded that “\[n\]egative social comparison . . . is a significant problem on Instagram, with 68% of teen girls experiencing negative social comparison on Instagram.” The report also concluded that the problem was not coming from celebrities or influencers but was rooted in the culture of Instagram itself. 209. Personalization Algorithm Feeding Harmful Content: In the summer of 2021, Meta internally distributed research showing that the personalization algorithm on Instagram Explore tends to increase “exposure to \[negative appearance comparison-provoking\] content beyond the preferences that people have indicated by the choice of accounts they follow.” Consequently, “17% of people see substantially more (at least 20 percentage points) \[negative appearance comparison-provoking\] content in Explore than in Feed. It’s worse for women and teen girls.” The study also concluded that the personalization algorithm showed users, in particular women and girls, significantly more “High Negative Appearance” content than the user chose to see on their feed. 210. On August 27, 2021, an Instagram spokesperson shared an exchange with Adam Mosseri about a forthcoming article by The Wall Street Journal’s Jeff Horwitz “that essentially argues that \[Instagram’s\] design is inherently bad for teenage girls (leads to \[suicide and self-harm\], poor mental health, dysphoria).” The spokesperson observed that Horwitz’s “arguments \[are\] based on \[Meta’s\] own research so \[they\] are difficult to rebut” and noted that the article could expose “that \[Meta’s\] own research confirmed what everyone has long suspected.” ## G. Meta publishes safety data it knows is misleading. 211. Meta regularly publishes public Community Standards Enforcement Reports (“the Reports” or “CSER”) that boast very low rates of its community standards being violated, while omitting from those reports Meta’s internal user experience survey data showing high rates of users’ actual encounters with harmful content. 212. Contrary to what Meta represents to the public, including in these Reports, Meta’s internal surveys frequently and consistently find that teens are exposed to harmful content and experience harmful interactions at higher rates than the general user population. In many instances, teen girls experience these harms, including inappropriate interactions from adults (such as solicitation or requests for nude photos), at higher rates than boys. 213. The Reports, published quarterly, describe the percentage of content posted on Instagram that Meta removes for violating Instagram’s community standards. Meta often refers to that percentage as its “Prevalence” metric. Through Report-related talking points, Meta directed its employees to tout the “Prevalence” metric as “the most important measure of a healthy online community.” 214. Meta has publicly represented that the “Prevalence” statistics in the Reports are a reliable measure of the safety of its Social Media Platforms, even going so far as to assert that the CSER “Prevalence” numbers were “the internet’s equivalent” of scientific measurements utilized by environmental regulators to assess the levels of harmful pollutants in the air. 215. The Reports are used by Meta to implicitly represent that, because Meta aggressively enforces its community standards—thereby reducing the “Prevalence” of community-standards-violating content—Meta’s Social Media Platforms are safe products that only rarely expose users (including children) to harmful content and harmful experiences. However, that representation is false. 216. Meta understands the disparity between Meta’s CSER Reports that it makes public and its internal survey results, but it continues to represent the misleading CSER Reports as accurate. 217. Meta uses several surveys to measure user experiences, including child and teen users, with their products, including exposure to harmful content and experiences. Two of these surveys include the Tracking Reach of Integrity Problems Survey (“TRIPS”) and the Bad Experiences & Encounters Framework (“BEEF”). 218. Contrary to the CSER data presented to the public, the TRIPS and BEEF data show much higher exposure to harmful content. 219. For example, Meta’s third quarter 2021 CSER Report stated that on Instagram, “less than 0.05% of views were of content that violated our standards against suicide & self-injury.” That representation creates the impression that it is very rare for users to experience content relating to suicide and self-injury on Instagram. 220. But Meta’s contemporaneous internal TRIPS and BEEF survey data showed that during 2021, 6.7% of surveyed Instagram users had seen self-harm content within the last seven days. For users between thirteen and fifteen years of age, 16.9% had seen content relating to self-harm on Instagram within the last seven days. 221. In other words, whereas a reader of the CSER Reports could reasonably understand Meta to be reporting that self-harm content on Instagram is rarely encountered by users—far less than 1% of the time—in reality, Meta knows from its user experience surveys that Instagram users, and children in particular, encounter content related to self-harm on Instagram much more commonly. 222. Despite the purported importance and centrality of “Prevalence” to Meta, Meta did not have a Prevalence metric for bullying and harassment for several years and did not provide a bullying and harassment Prevalence metric in Reports cycles until 2021. 223. The third quarter 2021 Report stated that “we estimate between 0.05% to 0.06% of views were of content that violated our standards against bullying & harassment \[on\ \ \ * * *\ \ Instagram\].” This representation creates the impression that it is very rare for users to experience bullying or harassment on Instagram. 224. Again, Meta’s contemporaneous internal user survey data told a different story: among surveyed Instagram users, 28.3% witnessed bullying on the platform within the last seven days and 8.1% were the target of bullying on the platform within the last seven days. 225. Among thirteen- to-fifteen-year-olds, 48.7% reported witnessing bullying within seven days. Among users aged sixteen to seventeen, that figure was 50%. 226. Similarly, and contrary to the 2021 third-quarter Report’s representation that harassment on Instagram was rare, Meta’s contemporaneous internal survey showed that 11.9% of all survey respondents said they had received unwanted advances on Instagram within the last seven days. 227. Among thirteen- to-fifteen-year-olds, 24.4% reported that they had received unwanted advances within the last seven days. Among sixteen- and seventeen-year-olds, that figure was 25.4%. 228. It’s not just that Meta emphasizes the data that makes its Platforms appear safe: Meta actively hides the data that shows the prevalence of users experiencing SSI, bullying, and harassment. Meta knew exactly what it was doing. The District obtained testimony from Meta’s former Director of Integrity during its investigation. When asked if he believed “that Mr. Zuckerberg and other company leaders focused on the ‘prevalence’ metric because it created a distorted picture about the safety of Meta’s Platforms,” he testified “I do.” When asked if he thought “Mr. Zuckerberg’s public statements about prevalence created a misleading picture of the harmfulness of Meta’s Platforms,” he testified “I do.” And when asked if he was aware of any instances where the company, in his view, minimized (i.e., downplayed) the harms users were experiencing on Meta’s Platforms, he testified: Every time that a Company spokesperson in the context of harms quotes prevalence statistics I believe that is what they are doing, that they’re minimizing the harms that people are experiencing in the product. COUNT I ## VIOLATIONS OF THE CONSUMER PROTECTION PROCEDURES ACT (Unfair Acts or Practices) 229. The District realleges and incorporates by reference each of the allegations contained in the preceding paragraphs as if set forth fully herein. 230. The CPPA is a remedial statute that is to be broadly construed. It establishes an enforceable right to truthful information from merchants about consumer goods and services that are or would be purchased, leased, or received in the District. 231. The services that Meta provides consumers are for personal, household, or family purposes and, therefore, are consumer goods and services. 232. Meta, in the ordinary course of business, supplies consumer goods and services and, therefore, is a merchant under the CPPA. 233. Users of Meta’s Social Media Platforms receive consumer goods and services from Meta in the form of social networking services and, therefore, are consumers under the CPPA. 234. The CPPA prohibits unfair and deceptive trade practices in connection with the offer, sale, and supply of consumer goods and services. 235. The CPPA prohibits any person from engaging in unfair trade practices. 236. By engaging in the acts and practices alleged herein, Meta engages in unfair acts or practices affecting District consumers, in violation of D.C. Code § 28-3904. * * * 237. Meta’s acts and omissions, as alleged herein, constitute unfair acts or practices within the meaning of the CPPA, including because they cause substantial injury to District consumers that those consumers cannot reasonably avoid. 238. The injuries caused by Meta are not outweighed by countervailing benefits to consumers or to competition. 239. As described in detail herein, Meta has, and at all relevant times had, a thorough understanding of the harms suffered by children who use its Platforms and the role its Platforms play in exacerbating existing harms. Instead of taking measures to mitigate these damaging effects, Meta turns a blind eye to them and persists in exploiting children’s psychological vulnerabilities. Through its acts and omissions, Meta knowingly causes unnecessary and unjustified harm to children for its own financial gain. 240. Meta’s acts and omissions alleged herein also have caused and continue to cause consumers substantial injury that those consumers could not and cannot reasonably avoid. Children cannot reasonably avoid injuries resulting from Meta’s acts and omissions, including because Meta misrepresents and fails to disclose the dangerous nature of its Platforms and because Meta utilizes and has utilized addictive, engagement-inducing features, knowing that children are especially susceptible to those psychologically manipulative tactics. 241. Meta’s unfair acts and practices include its choice to target its Social Media Platforms to children while knowingly designing them to include features that it knows to be psychologically and physically harmful to children—including features known to promote compulsive, prolonged, and unhealthy use by children. 242. Meta’s unfair design choices include deploying social features that unfairly harm children independently of any actions taken by third-party users of its Platforms. For example, * * * Meta unfairly utilizes features such as infinite scroll, ephemeral content, autoplay, and disruptive alerts to extract additional time and attention from children whose developing brains are not equipped to resist those manipulative tactics. 243. Meta has designed, developed, and deployed disruptive audiovisual and vibrating notifications, alerts, and ephemeral content features in a way that exploits children’s psychological vulnerabilities and cultivated a sense of “fear of missing out” in order to induce children to spend more time on Meta’s Platforms than they would otherwise. 244. By algorithmically serving content to children, according to “variable reward schedules,” Meta manipulates dopamine releases in children who use their Platforms, unfairly inducing them to engage repeatedly with their products—much like a child gambling at a slot machine. 245. Thus, in numerous instances, Meta has engaged in unfair practices by taking actions to facilitate children’s addiction to and unhealthy use of its Platforms. Meta’s choices to expose children to the features described above, individually and in combination, are an unfair act and practice, including because doing so causes children substantial injury that they cannot reasonably avoid, especially in view of their psychological and developmental vulnerabilities. There are no countervailing benefits to consumers or to competition that outweigh these substantial injuries. 246. Each unfair act or practice engaged in by Meta as recited above constitutes a separate violation of the CPPA. 247. Meta continues to cash in on the addictive nature of its Platforms despite knowing the harm its Platforms cause to the District’s children. * * * 248. Meta’s violations present a continuing harm, and the unlawful acts and practices complained of here affect the public interest. COUNT II ## VIOLATIONS OF THE CONSUMER PROTECTION PROCEDURES ACT (Misrepresentations and Omissions) 249. The District realleges and incorporates by reference each of the allegations contained in the preceding paragraphs as if set forth fully herein. 250. Meta, in the ordinary course of business, supplies consumer goods and services and, therefore, are merchants under the CPPA. 251. Users of Meta’s Social Media Platforms receive consumer goods and services from Meta in the form of social networking services and, therefore, are consumers under the CPPA. 252. The CPPA prohibits unfair and deceptive trade practices in connection with the offer, sale, and supply of consumer goods and services. 253. By engaging in the acts and practices alleged in this Complaint, Meta engages in deceptive acts or practices affecting District consumers, including by making or causing to be made to District consumers, directly or indirectly, explicitly or by implication, misrepresentations as to material facts which have a tendency to mislead consumers, in violation of D.C. Code § 28-3904(e), and by failing to state material facts and/or using innuendo or ambiguity as to material facts, which have a tendency to mislead District consumers, in violation of D.C. Code §§ 28-3904(f) and (f-1) 254. In numerous instances in connection with the advertising, marketing, promotion, and other representations regarding their products, including through the actions described herein, Meta has made deceptive representations, directly or indirectly, expressly or by implication, with the intent that consumers rely on them, including but not limited to the following: (a) misrepresenting that Meta’s Social Media Platforms are not designed to hook children; (b) misrepresenting the appropriateness of its Platforms for children and obscuring or hiding known risks that children will be exposed to extreme and harmful content, including suicide and self-harm; (c) misleading parents and consumers about features it knows promote body dysmorphia and eating disorders in children; (d) misleading consumers about the harms children face from social comparison on its Social Media Platforms; and (e) misrepresenting and downplaying its internal research findings about children and mental health. 255. In addition to Meta’s misleading statements, Meta’s omissions, which are false and misleading in their own right, render even seemingly truthful statements about Meta’s Social Media Platforms false and misleading. 256. Each deceptive act or practice engaged in by Meta as recited above constitutes a separate violation of the CPPA. 257. Meta’s unlawful acts and practices in violation of the CPPA target and adversely affect District residents. 258. Meta’s violations present a continuing harm, and the unlawful acts and practices complained of here affect the public interest. ## PRAYER FOR RELIEF WHEREFORE, the District of Columbia respectfully requests this Court enter a judgment in its favor and grant relief against Defendants as follows: (a) Issue, in accordance with D.C. Code § 28-3909(a), a permanent injunction that prohibits Defendants from engaging in unfair and deceptive conduct that harms District consumers, including District children, and from making false or misleading statements regarding their Social Media Platforms, in violation of the CPPA; (b) Order Defendants to pay restitution or damages in accordance with D.C. Code § 28-3909; (c) Award civil penalties in an amount to be proved at trial, and as authorized per violation of the CPPA, in accordance with D.C. Code § 28-3909(b); (d) Award the District the costs of this action and reasonable attorney’s fees in accordance with D.C. Code § 28-3909(b); and (e) Grant such further relief as the Court deems just and proper. ## JURY DEMAND The District of Columbia demands a jury trial by the maximum number of jurors permitted by law. Dated: October 24, 2023 Respectfully submitted, BRIAN L. SCHWALB Attorney General for the District of Columbia JENNIFER C. JONES Deputy Attorney General Public Advocacy Division BETH MELLEN Assistant Deputy Attorney General Public Advocacy Division /s/ Adam R. Teitelbaum ADAM R. TEITELBAUM (#1015715) Director, Office of Consumer Protection KEVIN VERMILLION (#1739318) Deputy Director, Office of Consumer Protection JORGE A. BONILLA LOPEZ (#1632867) Assistant Attorney General LYDIA E. MENDEZ (#90005211) * * * Assistant Attorney General Office of the Attorney General for the District of Columbia 400 6th St. NW, 10th Floor Washington, D.C. 20001 Tel: (202) 741-5226 [adam.teitelbaum@dc.gov](mailto:adam.teitelbaum@dc.gov) [kevin.vermillion@dc.gov](mailto:kevin.vermillion@dc.gov) [jorge.bonillalopez@dc.gov](mailto:jorge.bonillalopez@dc.gov) [lydia.mendez@dc.gov](mailto:lydia.mendez@dc.gov) Attorneys for the District of Columbia