`
`
`
`BERNSTEIN LITOWITZ BERGER
` & GROSSMANN LLP
`JONATHAN D. USLANER (Bar No. 256898)
`(jonathanu@blbglaw.com)
`CAITLIN BOZMAN (Bar No. 343721)
`(caitlin.bozman@blbglaw.com)
`2121 Avenue of the Stars, Suite 2575
`Los Angeles, CA 90067
`Tel: (310) 819-3470
`
`Counsel for Lead Plaintiffs
`Ohio Public Employees Retirement System and
`PFA Pension Forsikringsaktieselskab
`
`[Additional Counsel Appear on
`Signature Page]
`
`
`
`UNITED STATES DISTRICT COURT
`NORTHERN DISTRICT OF CALIFORNIA
`SAN JOSE DIVISION
`
`IN RE META PLATFORMS, INC.
`SECURITIES LITIGATION
`
`
`
`
`Lead Case No. 4:21-cv-08812-JST
`
`Consolidated Case No. 4:21-cv-08873-JST
`
`LEAD PLAINTIFFS’ CONSOLIDATED
`AMENDED CLASS ACTION
`COMPLAINT FOR VIOLATIONS OF
`THE FEDERAL SECURITIES LAWS
`
`CLASS ACTION
`
`
`
`
`
`
`
`
`
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`CONSOLIDATED AMENDED CLASS ACTION COMPLAINT
`4:21-cv-08812-JST
`
`
`
`
`
`
`Case 4:21-cv-08812-JST Document 97 Filed 10/28/22 Page 2 of 200
`
`
`
`I.
`
`II.
`
`TABLE OF CONTENTS
`
`Page
`
`INTRODUCTION ...............................................................................................................1
`
`JURISDICTION AND VENUE ........................................................................................15
`
`III.
`
`PARTIES ...........................................................................................................................16
`
`
`
`
`
`Lead Plaintiffs ........................................................................................................16
`
`Defendants .............................................................................................................16
`
`IV.
`
`SUMMARY OF THE FRAUD .........................................................................................18
`
`
`
`
`
`Before the Class Period, Meta Notices Material Negative Trends in Its User
`Base and Engagement that Threaten Its Profitability ............................................18
`
`Defendants Make a Series of Misstatements and Omissions Concerning
`Meta’s Content-Moderation Practices, Algorithm, the Impact of its
`Platforms on Its Users, and the Size of Its User Base ............................................23
`
`1.
`
`Contrary to Defendants’ Public Representations, Meta Exempted
`Several Million High-Profile Users from Its Content-Moderation
`Policies, Allowing Them to Violate “Community Standards” with
`Impunity .....................................................................................................23
`
`a.
`
`b.
`
`c.
`
`Defendants Assure Investors that Meta Applies Its
`Content Standards Equally to All Users ........................................24
`
`In Truth, Meta’s “Cross-Check” and “Whitelisting”
`Practices Exempted Several Million High-Profile
`Users from Content Enforcement ..................................................29
`
`the
`Meta’s Own Oversight Board Excoriated
`Company for Misrepresenting and Concealing
`Cross-Check ...................................................................................34
`
`2.
`
`Defendants Made Material Misstatements and Omissions
`Concerning the Fact That Meta’s News Feed Algorithm and
`Content-Moderation Practices Promoted Toxic Content and
`Misinformation ..........................................................................................37
`
`a.
`
`b.
`
`c.
`
`The News Feed Is Facebook’s Primary Content-
`Delivery Mechanism ......................................................................37
`
`Defendants Changed the News Feed Algorithm in
`2018, Stating
`that
`the Change Prioritized
`“Meaningful Social Interactions” ..................................................38
`
`Meta Quickly Determines that its 2018 Changes to
`the News Feed Algorithm Caused Hateful, Toxic
`Content and Misinformation to Be “Inordinately
`Prevalent” .......................................................................................39
`
`CONSOLIDATED AMENDED CLASS ACTION COMPLAINT
`4:21-cv-08812-JST
`
`
`
`i
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`Case 4:21-cv-08812-JST Document 97 Filed 10/28/22 Page 3 of 200
`
`d.
`
`e.
`
`f.
`
`Defendants Made Numerous False and Misleading
`Representations Concerning
`the News Feed
`Algorithm and Meta’s Content Moderation ...................................42
`
`Copious Internal Research Documented that the
`Algorithm Caused Toxic Content to Proliferate, But
`Meta and Zuckerberg Chose Not to Fix the Problem ....................47
`
`to
`and Unwillingness
`Inability
`Defendants’
`Effectively Moderate Content Made It Impossible to
`Achieve Meta’s Publicly Stated Objectives to
`Support COVID-19 Vaccination....................................................59
`
`3.
`
`Meta Falsely Touts Its Commitment to Children’s Safety and
`Downplays and Denies the Harms Caused by Its Platforms, Despite
`Knowing from Extensive Internal Research that Its Platforms
`Severely Harm Teens, Especially Teen Girls ............................................61
`
`a.
`
`b.
`
`c.
`
`d.
`
`e.
`
`f.
`
`Defendants Launch a Campaign to “Race for Teen
`Users” .............................................................................................62
`
`that
`Internal Research Demonstrates
`Meta’s
`Instagram’s Negative Impacts on Teens, Especially
`Teen Girls, Are “Common” And “Severe” ....................................64
`
`Defendants Conceal Meta’s Internal Research on
`Teen Mental Health from the Public ..............................................69
`
`“Winning with Tweens”: Meta Works to Attract
`Even Younger Users Because It Is Losing the “Race
`for Teens”.......................................................................................71
`
`Defendants Falsely Tout Meta’s Commitment to
`Children’s Safety, Assure the Public that Meta’s
`Platforms Do Not Harm Children, and Characterize
`the Research as Inconclusive .........................................................72
`
`Meta Doubles Down on Its Efforts to Target Young
`Users, Its False Assurances, and Its Denials of the
`Negative Impact of Its Platforms on Children ...............................73
`
`4.
`
`Meta Mispresents Its User Growth Metrics Due to the Prevalence of
`Duplicate or SUMA Accounts Among New Accounts .............................77
`
`a.
`
`b.
`
`c.
`
`d.
`
`User Growth Was a Key Metric for Investors ...............................77
`
`Meta Sold Facebook as a “Real Identity” Platform
`Where Advertisers Could Reliably Target Key
`Demographics like Teens and Young Adults ................................78
`
`Meta Misstates Facebook’s User Growth Metrics .........................79
`
`In Truth, Meta Knew that Duplicate Accounts Were
`Far More Prevalent in New Accounts than Reported,
`and Teen and Young Adult Use and Engagement
`Was Actually in Decline ................................................................81
`
`CONSOLIDATED AMENDED CLASS ACTION COMPLAINT
`4:21-cv-08812-JST
`
`
`
`ii
`
`
`
`Case 4:21-cv-08812-JST Document 97 Filed 10/28/22 Page 4 of 200
`
`
`
`Meta Stock Price Declines in Response to the Publication of the Facebook
`Files and Additional Disclosures Connected to the Fraud .....................................83
`
`1.
`
`The Wall Street Journal Publishes the Facebook Files and
`Defendants Push Back with Denials ..........................................................83
`
`a.
`
`b.
`
`c.
`
`September 13, 2021: the September 13 Facebook
`File Describes Internal Documents Showing that
`Meta Allowed Millions of Users to Violate Its
`Community Standards and Post Harmful Content
`with Impunity .................................................................................85
`
`September 14, 2021: the September 14 Facebook
`File Showcases How Meta Profited by Making
`Instagram “Toxic”
`for Teens, Causing New
`Congressional Scrutiny ..................................................................88
`
`September 15, 2021: the September 15 Facebook
`File Describes How Meta’s Changes to Its News
`Feed Algorithm and Content Moderation Failures
`Generated Prevalent Toxic Content ...............................................96
`
`2.
`
`Meta’s Stock Price Declines Further in Response to Subsequent
`Disclosures Connected to the Fraud ..........................................................99
`
`a.
`
`b.
`
`c.
`
`d.
`
`e.
`
`September 21-22, 2021: Meta’s Oversight Board
`Calls for More “Transparency” on X-Check, While
`Senators Grill Meta over How Facebook Harms
`Teens ..............................................................................................99
`
`September 27-28, 2021: Meta Pauses Instagram
`Kids, The Wall Street Journal Reveals How Meta
`Planned to Attract Preteens to Its Platforms, and
`Meta Seeks a Review of X-Check by the Oversight
`Board ............................................................................................101
`
`October 3, 2021: the Facebook Files Whistleblower
`Is Revealed ...................................................................................105
`
`October 6, 2021: Meta Halts New Product Launches
`and European Regulators Work with
`the
`Whistleblower ..............................................................................110
`
`October 21, 2021: the Oversight Board Condemns
`Meta for Withholding Relevant Information About
`X-Check, Meta Admits that Its Statements to the
`Oversight Board Were Misleading, and The Wall
`Street Journal Releases New Information About
`Meta’s User Metrics .....................................................................112
`
`(1)
`
`(2)
`
`The Oversight Board Formally Reviews X-Check ..........112
`
`The Wall Street Journal Reveals That Up to 56% of
`New Accounts Are Created by Existing Users ................113
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`CONSOLIDATED AMENDED CLASS ACTION COMPLAINT
`4:21-cv-08812-JST
`
`
`
`iii
`
`
`
`Case 4:21-cv-08812-JST Document 97 Filed 10/28/22 Page 5 of 200
`
`
`
`V.
`
`POST-CLASS PERIOD DEVELOPMENTS ..................................................................115
`
`
`
`
`
`
`
`
`
`Another Whistleblower Corroborates Haugen’s Complaints and the
`Facebook Files .....................................................................................................115
`
`A News Outlet Consortium Gains Access to the Facebook Papers,
`Corroborating the Facebook Files and Expanding on the Platforms’ Harms ......115
`
`Analysts React to the Facebook Files and Ensuing Developments .....................116
`
`Investigations Concerning Meta’s Promotion of Instagram to Children and
`Young Adults Commence as Scrutiny Intensifies ...............................................117
`
`VI.
`
`DEFENDANTS’ MATERIALLY FALSE AND MISLEADING STATEMENTS
`AND OMISSIONS ..........................................................................................................119
`
`
`
`
`
`
`
`
`
`False and Misleading Statements Representing that Meta Applied Its
`Content-Moderation Policies Consistently to All Users ......................................119
`
`False and Misleading Statements Concerning Presence and Impact of
`Misinformation and Harmful Content on Meta’s Platforms ................................132
`
`1.
`
`2.
`
`Defendants’ False and Misleading Statements Concerning Meta’s
`Algorithms and Harmful Content on Meta’s Platforms ...........................132
`
`Additional False and Misleading Statements Concerning Meta’s
`Ability and Willingness to Moderate Misinformation and Harmful
`Content on Its Platforms ..........................................................................141
`
`False and Misleading Statements Denying Instagram’s Harm to Young
`Users ....................................................................................................................150
`
`False and Misleading Statements Representing that Meta’s Metrics for
`Counting Users Were Accurate and Reliable ......................................................160
`
`VII. ADDITIONAL LOSS CAUSATION ALLEGATIONS .................................................163
`
`VIII. ADDITIONAL SCIENTER ALLEGATIONS ................................................................166
`
`IX.
`
`INAPPLICABILITY OF STATUTORY SAFE HARBOR ............................................186
`
`X.
`
`THE PRESUMPTION OF RELIANCE ..........................................................................187
`
`XI.
`
`CLASS ACTION ALLEGATIONS ................................................................................188
`
`XII. CLAIMS FOR RELIEF ...................................................................................................190
`
`COUNT I .........................................................................................................................190
`
`COUNT II ........................................................................................................................192
`
`XIII. PRAYER FOR RELIEF ..................................................................................................193
`
`CONSOLIDATED AMENDED CLASS ACTION COMPLAINT
`4:21-cv-08812-JST
`
`
`
`iv
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`
`
`Case 4:21-cv-08812-JST Document 97 Filed 10/28/22 Page 6 of 200
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`Lead Plaintiffs Ohio Public Employees Retirement System (“OPERS”) and PFA Pension,
`
`Forsikringsaktieselskab (“PFA Pension”) (together, “Lead Plaintiffs”), by and through their
`
`counsel, bring this action under Sections 10(b) and 20(a) of the Securities Exchange Act of 1934
`
`(the “Exchange Act”) on behalf of themselves and all persons or entities, except Defendants and
`
`other persons and entities excluded below, who purchased or otherwise acquired Meta Platforms,
`
`Inc. f/k/a Facebook, Inc. (“Meta” or the “Company”) Class A common stock between April 27,
`
`2021, and October 21, 2021, inclusive (the “Class Period”), and were damaged thereby.
`
`Lead Plaintiffs’ information and belief as to allegations concerning matters other than
`
`themselves and their own acts is based upon the investigation conducted by and through counsel,
`
`which included, among other things, the review and analysis of (i) transcripts, press releases, news
`
`articles, and other public statements issued by or concerning Meta; (ii) research reports issued by
`
`financial analysts concerning the Company; (iii) reports and other documents filed publicly by
`
`Meta with the U.S. Securities and Exchange Commission (“SEC”); (iv) Meta’s corporate website;
`
`and (v) other material and data identified herein. Lead Counsel’s investigation into the factual
`
`allegations is continuing, and many of the relevant facts are known only by Defendants or are
`
`exclusively within their custody or control. Lead Plaintiffs believe that substantial additional
`
`evidentiary support will exist for these allegations after a reasonable opportunity for discovery.
`
`18
`
`I.
`
`INTRODUCTION
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`1.
`
`This case arises from misstatements and omissions by the senior executives of
`
`Meta—the most powerful social media company in the world—about the severe harms that its
`
`platforms cause, and its decision to prioritize profits over users’ safety. Several stunning
`
`disclosures in September and October 2021 brought these misrepresentations and omissions to
`
`light, outraged the public, and triggered a wave of investigations that remain ongoing.
`
`2.
`
`The disclosures began with the September 2021 publication of the “Facebook
`
`Files,” a series of bombshell articles by The Wall Street Journal that revealed previously secret
`
`internal Meta research and communications provided by an anonymous former Meta employee
`
`turned whistleblower. These articles started to rip the lid off what the whistleblower described as
`
`Meta’s “black box” concealing how serious these harms were, what Defendants knew about them,
`
`CONSOLIDATED AMENDED CLASS ACTION COMPLAINT
`4:21-cv-08812-JST
`
`
`
`1
`
`
`
`Case 4:21-cv-08812-JST Document 97 Filed 10/28/22 Page 7 of 200
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`and how Meta repeatedly allowed these harms to continue to maximize revenue.
`
`3.
`
`The disclosures cascaded into October 2021 when Frances Haugen—a former Meta
`
`data scientist who studied these harms—revealed herself as the whistleblower who provided the
`
`materials to the Journal. Haugen gave a shocking accounting of what she had witnessed at Meta,
`
`first on 60 Minutes and in an interview with the Journal, and later in sworn testimony to Congress.
`
`Haugen explained that the final straw for her was Meta’s December 2, 2020, decision to disband
`
`its Civic Integrity Unit—the unit charged with policing how the platform could spread falsehoods,
`
`stoke violence, and disseminate toxic content.
`
`4.
`
`Rather than permit Meta to continue to escape accountability for the decisions it
`
`was making behind closed doors, Haugen gathered thousands of pages of internal Meta research
`
`before she left the Company. Haugen then provided this research to the Journal, Congress, and the
`
`SEC to expose the brutal truth that Meta, “over and over again, has shown it chooses profit over
`
`safety. It is subsidizing . . . its profits with our safety.”
`
`5.
`
`Meta is the world’s largest social media company, and owns and operates the
`
`ubiquitous Facebook and Instagram platforms. Meta makes essentially all its revenue—which
`
`totaled a staggering $115 billion in 2021—from advertising. Meta’s advertising revenue depends
`
`on two fundamental metrics: the size of its user base and users’ level of engagement, i.e., how
`
`active they are on the platform. As Meta told investors, “The size of our user base and our users’
`
`level of engagement are critical to our success,” as its advertising revenue was highly dependent
`
`on its ability to “continue to increase user access to and engagement with our products.”
`
`6.
`
`In the years before the Class Period, Meta’s executives observed concerning
`
`downward trends in user engagement and growth—trends that, if not arrested and reversed, would
`
`have a substantial negative impact on Meta’s revenue stream. Meta’s users were becoming less
`
`engaged with its platforms. As reported by Forbes and The Wall Street Journal, users “posted a
`
`third less content in 2016 than in 2015,” and Meta’s training videos and internal memos
`
`acknowledged that “engagement was broadly declining.” This, in turn, impacted the performance
`
`of Meta’s advertisements significantly. For example, a 2018 study conducted by two social media
`
`consulting companies found that the “top 20,000 brands on Facebook . . . suffered more than a 50
`
`CONSOLIDATED AMENDED CLASS ACTION COMPLAINT
`4:21-cv-08812-JST
`
`
`
`2
`
`
`
`Case 4:21-cv-08812-JST Document 97 Filed 10/28/22 Page 8 of 200
`
`
`
`percent reduction in engagement.”
`
`7.
`
`Meanwhile, Meta struggled mightily to penetrate the teen, young adult, and
`
`“tween” markets. This age cohort is particularly important to social media companies because
`
`young users create trends, adopt new technologies, and represent a critical untapped source for
`
`new growth. Internal Meta research dated February 20, 2020, and titled, “What we know about the
`
`continued decline in OGPS,” noted that young adults were “less engaged on Facebook than older
`
`adults.” An internal Meta presentation from November 2020, titled, “Young Adults FB Strategy,”
`
`reaffirmed this decline in young adult engagement and described it as “a significant risk” to the
`
`user and engagement metrics that Meta reported publicly and relied on to drive revenue. Perhaps
`
`more alarming, teens were even more apathetic toward Meta’s products than young adults; Meta’s
`
`own study, entitled “Future of YA,” indicated that from 2019 to 2021, the number of active teenage
`
`users declined sharply, and “sessions, time spent, [and] production [were] all much worse among
`
`teens than young adults.” These negative trends threatened the Company’s ability to charge
`
`advertisers a premium for space on Meta’s platforms.
`
`8.
`
`In response, Meta undertook several initiatives designed to boost engagement,
`
`attract users (particularly young users) and, ultimately, enhance advertising revenue. For example,
`
`in 2018, Meta made a fundamental change to its central algorithm—the formula that determines
`
`which pieces of content Facebook users view in their News Feed—to increase engagement. Meta
`
`also promulgated and purported to rigorously enforce Community Standards, a set of rules about
`
`what content was permitted on Facebook, in order to control the spread of harmful content and
`
`thus encourage users to engage more. To penetrate the teen and young adult market, Meta heavily
`
`invested in Instagram and developed plans to roll out initiatives designed to leverage that platform
`
`to attract new users as young as six years old. Meta also promoted itself as a “real identity”
`
`platform, i.e., a platform largely free from fake or duplicate accounts, and said that the prevalence
`
`of single users with multiple accounts (or “SUMA”) was relatively low at 11%.
`
`9.
`
`By outward measures, these initiatives worked. Meta’s engagement increased, and
`
`its revenues sharply rose—more than doubling from $40.7 billion in 2017 to $86 billion in 2020.
`
`Meta’s stock price also skyrocketed, rising from $117 to $273 per share in that timeframe.
`
`CONSOLIDATED AMENDED CLASS ACTION COMPLAINT
`4:21-cv-08812-JST
`
`
`
`3
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`
`
`Case 4:21-cv-08812-JST Document 97 Filed 10/28/22 Page 9 of 200
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`10.
`
`As the Class Period approached, however, Meta found itself at the center of
`
`controversy stemming from the platform’s role in the political discord and violence of January 6,
`
`2021. Meta came under intense scrutiny from Congress, regulators, and the investing public, who
`
`questioned whether the Company was an instrument of disinformation, toxic content, and other
`
`harmful impacts on users. On March 25, 2021, Defendant Zuckerberg was brought to testify before
`
`Congress concerning, among other things, the Company’s role in spreading misinformation and
`
`whether its platforms harmed children. One securities analyst noted that the hearing was part of a
`
`“regulatory onslaught” that could force content-moderation changes and additional government
`
`regulation—including a break-up of the powerful Company.
`
`11.
`
`These issues were enormously important to Meta, its revenue stream and growth
`
`potential, and its investors. For instance, if it were true that Meta consciously allowed
`
`misinformation or toxic content to proliferate on its platform, or knowingly allowed its products
`
`to harm young users, this would irreparably damage Meta’s ability to attract and engage users
`
`across the board.
`
`12.
`
`Accordingly, throughout the Class Period, Defendants made a series of statements
`
`on multiple subjects designed to assuage concern, deflect the scrutiny Meta faced, and clear the
`
`way for the Company to continue its efforts to boost engagement, attract users, and reap enormous
`
`profit. These statements propelled Meta’s stock price to a Class Period—and all-time—high of
`
`over $384 per share on September 1, 2021. As the facts revealed by Haugen would soon show,
`
`however, these statements were materially false and misleading.
`
`13. Misstatements and Omissions Concerning the “X-Check” and “Whitelisting”
`
`Program, Through Which Meta Exempted Millions of Users from Its Content Policies. Meta
`
`repeatedly assured the investing public that it applied its Community Standards equally to all users,
`
`regardless of the user’s status. Meta purportedly “recognizes how important it is for Facebook to
`
`be a place where people feel empowered to communicate, and we take our role seriously in keeping
`
`abuse off the service.” The equal application of Meta’s Community Standards was key to attracting
`
`new users and supporting engagement: users will join and engage with a social media platform
`
`only if they feel that the platform does not permit harmful content and misinformation to
`
`CONSOLIDATED AMENDED CLASS ACTION COMPLAINT
`4:21-cv-08812-JST
`
`
`
`4
`
`
`
`Case 4:21-cv-08812-JST Document 97 Filed 10/28/22 Page 10 of 200
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`proliferate, and that the platform applies its content rules equally to all users regardless of their
`
`celebrity, politics, or other status.
`
`14.
`
`For precisely these reasons, Meta stated unequivocally, “Our Community
`
`Standards apply to all content, and we assess everyone under those Standards,” and “we remove
`
`content from Facebook, no matter who posts it, when it violates our Community Standards.”
`
`Defendant Nick Clegg (Meta’s then-Vice President of Global Affairs) likewise emphasized that
`
`“whether you are the Pope, the Queen, or the President of the United States, what we apply to
`
`everybody is you cannot use our services to say things which we think can deliberately lead to
`
`harm.”
`
`15. Meta had an Oversight Board, an independent body that the Company formed to
`
`supposedly review Meta’s difficult content-moderation decisions. Meta stated to this Board (and
`
`the public) that Meta reviewed certain content decisions using a process that it called X-Check or
`
`“cross-check.” Meta stated that the review under cross-check was done in only “a small number
`
`of decisions.” Significantly, Meta further stated that the cross-check review was done only to
`
`ensure that the content-moderation decisions were correct under Community Standards, and thus
`
`“minimize the risk of errors in enforcement.” Meta further stated that cross-check did not change
`
`the fact that the Company applied its Community Standards “consistently and fairly.”
`
`16.
`
`In truth, Meta did not “apply [its Community Standards] to all content,” did not
`
`“remove content from Facebook, no matter who posts it, when it violates our Community
`
`Standards,” and did not treat “the Pope, the Queen, or the President of the United States” like it
`
`treated the average user. Further, cross-check was certainly not a mere second-level review used
`
`only to “minimize the risk of errors in enforcement” of the Community Standards. Unknown to
`
`investors, cross-check was an elaborate mechanism by which Meta simply exempted a massive
`
`amount of users—including its most influential users—from its Community Standards, thus
`
`allowing them to post content that violated standards with impunity. As investors would learn at
`
`the end of the Class Period when The Wall Street Journal published internal Meta research and
`
`communications, cross-check “shield[ed] millions of VIP users from the company’s normal
`
`enforcement process,” including “whitelisted” users, who were “rendered immune from
`
`CONSOLIDATED AMENDED CLASS ACTION COMPLAINT
`4:21-cv-08812-JST
`
`
`
`5
`
`
`
`Case 4:21-cv-08812-JST Document 97 Filed 10/28/22 Page 11 of 200
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`enforcement actions,” and others who were “allowed to post rule-violating material pending
`
`Facebook employee reviews that often never came.”
`
`17.
`
`The scope of this program was enormous: Meta maintained 45 global teams that
`
`kept and monitored the exempt list. The number of users protected by the program had grown to
`
`nearly 6 million prior to the Class Period. The exempted elite included many of the most prominent
`
`people in the world, whose content generated a disproportionate amount of views and had an
`
`outsized impact on public discourse. Meta’s undisclosed program allowed rule-violating content
`
`to be viewed more than one hundred billion times.
`
`18.
`
`As reflected in an internal Meta research report titled “Comparing the effects of
`
`misinfo from politicians vs ordinary user sources,” Meta knew that this whitelisted content had a
`
`uniquely harmful effect. Meta’s research showed that “[u]nder the political whitelist policy,
`
`content posted by whitelisted Pages or Profiles is not subject to our misinformation interventions,”
`
`and “results indicate the current whitelist policy is protecting content that is especially likely to
`
`deceive, and hence to harm, people on our platforms.” Accordingly, the report “recommend[ed]
`
`ending the whitelist, or at least (1) informing users; [and] (2) reducing distribution”—none of
`
`which Meta did.
`
`19.
`
`Given its size and the fact that it was the mechanism that permitted much of
`
`Facebook’s most controversial (and destructive) content to go viral, Meta’s senior executives were
`
`intimately familiar with the cross-check and whitelisting program. Indeed, Meta’s most senior
`
`executives personally participated in the program, often making key decisions about whom to
`
`include. An internal Meta September 2020 memo described what the Journal called “daily
`
`interventions in [the Company’s] rule-making and enforcement process by both Facebook’s
`
`public-policy team and senior executives.” Meta employees with responsibility for content
`
`moderation raised their concerns internally that “the final judgment about whether a prominent
`
`post violates a certain written policy are made by senior executives, sometimes Mark Zuckerberg,”
`
`but “[i]f our decisions are intended to be an application of a written policy then it’s unclear why
`
`executives would be consulted.”
`
`20. Meta employees internally recognized that the Company was lying about this
`
`CONSOLIDATED AMENDED CLASS ACTION COMPLAINT
`4:21-cv-08812-JST
`
`
`
`6
`
`
`
`Case 4:21-cv-08812-JST Document 97 Filed 10/28/22 Page 12 of 200
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`program publicly and that it raised serious ethical concerns. For instance, a 2019 internal Meta
`
`audit found that whitelist status, which was sometimes granted without any documented reason,
`
`was “pervasive, touching almost every area of the company,” and that the practice was “not
`
`publicly defensible.” As the audit stated, “We are not actually doing what we say we do publicly,”
`
`and whitelisting “pose[d] numerous legal, compliance, and legitimacy risks for the company and
`
`harm to our community.”
`
`21.
`
`Knowing that the revelation of these facts would damage its reputation and revenue
`
`base, Meta kept the true nature of the program a closely guarded secret from the public and even
`
`its own Oversight Board. As detailed herein, multiple times during the Class Period, the Oversight
`
`Board called on Meta to provide additional “transparency” about the nature of the cross-check
`
`program. Meta consistently provided vague, incomplete, and misleading answers. When the truth
`
`about the program was ultimately revealed toward the end of the Class Period (described further
`
`below), the Oversight Board excoriated Meta. The Oversight Board found that Meta “has not been
`
`fully forthcoming in its response on cross-check,” and instead gave “ambiguous, undetailed”
`
`responses that provided “no meaningful transparency” and were “not acceptable.” Tellingly,
`
`confronted with the information made public from the Facebook Files, Meta itself “admitted” to
`
`the Oversight Board that its statements “could come across as misleading.”
`
`22. Misstatements and Omissions Concerning the Fact that the News Feed
`
`Algorithm and Meta’s Content-Moderation Practices Caused Misinformation and Harmful
`
`Content to Proliferate. In 2018, Meta announced and implemented substantial changes to
`
`Facebook’s News Feed algorithm, fundamentally changing the process by which the Company
`
`determined what content users would view on the platform. Meta repeatedly stated to investors
`
`that those algorithm changes simply prioritized “Meaningful Social Interactions” (“MSI”), and did
`
`not cause harmful content—e