Scroll to read the report

How Social-Media Rollbacks Endanger Democracy
Ahead of the 2024 Elections

Big Tech Backslide

It’s imperative that platforms redouble their previous efforts to root out lies, hate and violence. Companies should:

Reinvest in and bolster staffing of teams needed to safeguard election integrity, trust and safety, and moderation.
Immediately reinstate protective policies to moderate election-related content and political ads, provide labeling transparency, enforce bans on COVID disinformation, and offer specific safeguards against targeted harassment.
Launch 2024 election-specific platform interventions in time for the U.S. primaries — and keep these protections in place through at least February 2025.
Hold VIP accounts to the same enforcement standards applied to other users.
Develop and implement more efficient review, labeling and enforcement against falsehoods in political ads across languages.
Develop improved transparency and disclosure practices, including regularly sharing core-metrics data with external researchers, journalists, lawmakers and the public. Provide quarterly reports on key trends, virality reports, network analysis and more.

TOPLINE RECOMMENDATIONS

Against the backdrop of these rollbacks, there is a looming new threat to platform integrity: Elon Musk has filed lawsuits against independent researchers in an attempt to silence criticism and evade accountability.⁸

There are dangerous real-world consequences when companies retreat from previous commitments to platform integrity, content moderation and robust enforcement of their terms of service.⁹ Insurrectionists challenged the results of the 2020 U.S. presidential election in part due to conspiracy theories that the platforms amplified.¹⁰ The entities overseeing 13 presidential libraries — Democratic and Republican alike — warned in September that U.S. democracy is in a fragile state, with civility and public discourse under grave threat.¹¹ And the platforms’ failures to curb hate and disinformation related to the Israel-Hamas war has fueled mass violence.¹²

Whether it’s in times of crisis or calm, social media has typically been a source of real-time information for users. And people should be able to rely on the platforms they use to provide accurate and even lifesaving information. The failure to vet and remove violative content harms and alienates users. Failure to moderate content inevitably leads to migration of platform lies and toxicity to mainstream media.¹³

Since the Jan. 6 insurrection, other real-world crises — like the attempted coup in Brazil in January 2023 and the conflict in the Middle East — have illustrated the critical role social-media platforms play in shaping rapidly unfolding events. Over and over we’ve seen how people can weaponize social media to sow division, undermine democracy and even fuel calls for violence offline.¹⁴ Without the policies and teams they need to moderate violative content, platforms risk amplifying confusion, discouraging voter engagement and creating opportunities for network manipulation to erode democratic institutions.

Litigation to Chill Tech Accountability

Mass layoffs across critical teams signal that these platforms are deemphasizing content moderation and user safety:

Meta, Twitter and YouTube have laid off approximately 40,750 employees across the three companies.⁷ Significant cuts occurred in the trust and safety, ethical engineering or responsible innovation, and content-moderation/consultant categories — the teams otherwise tasked with maintaining the platforms’ general health.

Mass Layoffs Make Content Moderation Harder

$\"Photo$

Meta, Twitter and YouTube — the three biggest platforms — have rolled back crucial policies that had previously kept hate, harassment and lies in check. This waning commitment to content moderation has led to a spike in hate and disinformation, eroding people’s experiences of these products and posing a risk to user safety.

Between Nov. 1, 2022 and Nov. 1, 2023, Meta, Twitter² and YouTube eliminated a total of 17 critical policies across their platforms:

Twitter and YouTube rolled back election-misinformation policies designed to limit “Big Lie” content about the 2020 election.
Twitter and Meta rolled back policies that had banned COVID-19 disinformation.
Twitter began allowing political ads on the platform. Meta stopped applying a political-ad policy that had mandated transparency and labeling in such advertisements.³ Both policies will allow for heightened disinformation in the ads users see on these platforms.
Meta and Twitter have both weakened privacy protections for users to allow for greater use of people’s data for the training of AI tools. Google has started giving its Bard AI tool access to YouTube, raising questions about how this chatbot will exploit user data.⁴
Meta, Twitter and YouTube reinstated Donald Trump’s accounts despite his outsized role in supporting and fueling the Jan. 6 insurrection.⁵ Meta and YouTube have refused to apply their policies as stringently to Trump as they do to everyday users.
Notably, TikTok has not rolled back any of its policies; in fact, in some instances it has strengthened platform features and policies. Yet TikTok remains unprepared to responsibly moderate toxic content on its platform. It enforces its policies in a lackluster manner and, at times, has downranked activists’ posts that do not appear to violate its policies.⁶

Policy Rollbacks Deprioritize User Safety & Platform Integrity

$\"\"/$

After the deadly insurrection at the U.S. Capitol on Jan. 6, 2021, tech companies finally seemed to accept that their failure to moderate content was undermining public safety and democracy. In the aftermath, most social-media companies removed users who spread anti-democratic conspiracies or used their online platforms to incite violence. Leading up to the 2022 U.S. midterm elections, tech companies promised users, civil-society groups and governments that they would safeguard election integrity and free expression on their platforms.

In 2022, Free Press graded the four major platforms’ policies against 15 recommendations we helped develop that are designed to curb the spread of election disinformation and extremism. Our research found that although tech companies have long promised to fight disinformation and hate, there is a notable gap between what the companies say they want to do and what they actually do in practice. Companies like Meta, TikTok, Twitter and YouTube do not have sufficient policies, practices, AI systems or human capital in place to materially mitigate harm ahead of, during and after election periods.¹

These failures have only mushroomed since we released our report. We found that in 2023, the largest social-media companies have deprioritized content moderation and other user trust and safety protections, including rolling back platform policies that had reduced the presence of hate, harassment and lies on their networks. These companies have also laid off critical staff and teams tasked with maintaining platform integrity. Taken together with the preferential treatment of VIP users — reflected in the reinstatement of Donald Trump’s accounts on Meta, Twitter and YouTube — these developments represent a dangerous backslide. In turn, this has created a toxic online environment that is vulnerable to exploitation from anti-democracy forces, white supremacists and other bad actors.

Executive Summary

Elections aren’t happening just in the United States in 2024, when 40 national elections will occur worldwide.¹⁵

Major social-media companies have long failed when it comes to applying corporate policies robustly and equitably across the globe.¹⁶ Platform executives have met civil society’s requests for improvement with indignation, denial and inaction. Congressional hearings to investigate possible liability have yielded little insight into platform business practices.¹⁷ Letters from legislators and dignitaries around the world asking for data on their algorithms, enforcement and staffing have resulted in few meaningful disclosures.¹⁸

For years, dozens of public-interest groups like Free Press have offered recommendations and expert guidance to the major social-media platforms. But these companies have responded with inconsistent and often lackluster commitments to reform.¹⁹ However, the persistent pressure from organizations and our civil-rights coalitions — such as Change the Terms and #StopToxicTwitter, both of which Free Press helps lead — has resulted in some powerful wins:

Strengthening Meta’s Dangerous Organizations and Individuals Policy: Meta, then Facebook, has made several changes to its policies in response to pressure from the Change the Terms coalition. In 2019, Facebook changed its dangerous-organizations policy to include white nationalists. In 2020 — after Free Press and allies in the Stop Hate for Profit coalition organized an advertising boycott — the platform added QAnon under its dangerous-organizations policy and began enforcing its policy to remove violent, extremist QAnon content.²⁰
Prioritizing Equity in Meta’s Moderation Across Languages: For years, Free Press and allies pushed social-media platforms to invest significant resources in combating hate and disinformation in languages other than English. After Free Press launched the #YaBastaFacebook campaign, and whistleblower Frances Haugen revealed failures in Meta’s enforcement of non-English content, the company finally committed to fighting misinformation across all languages. We’re continuing to push Meta and other platforms to enforce their policies equitably and to protect non-English users.
Providing Transparent Access for Researchers: In a big win for transparency, TikTok announced in 2022 that it will allow researchers to delve into its data, evaluate its content and test its moderation system. To prove that it operates independently of the Chinese government, the company also announced that it will allow Oracle to audit its algorithms and content-moderation models. In another win, YouTube invited researchers to apply for access to its global data.
Advancing Platform Integrity at Pre-Musk Twitter: Following meetings with Change the Terms leaders in 2018, Twitter banned deadnaming and misgendering.²¹ In 2019, following our #StopRacistTwitter initiative and public protest outside the company’s headquarters, Twitter bolstered its hateful-conduct policies to rein in violent and deceitful language. The company also banned political advertising, citing support from civil society and the platform’s Trust and Safety Council.²² In 2020, it expanded its disinformation policy to include COVID disinformation and provided a clear label for misleading content.²³
Preserving Election-Integrity Measures on Twitter Following Musk’s Purchase: Free Press Co-CEO Jessica J. González and close allies met directly with Musk soon after he assumed ownership of the platform in October 2022 and got him to pledge that he wouldn’t reinstate banned accounts before the midterm elections. When Musk gutted content-moderation policies and laid off thousands of key employees, Free Press partnered with Accountable Tech, Color Of Change, Media Matters for America and dozens of other allies to launch the #StopToxicTwitter coalition, which helped push more than 50 percent of Twitter’s top-100 advertisers to pause their spending on the platform. These efforts slowed the pace of Musk’s Twitter rollbacks and reduced the chance that other platforms would weaken their own moderation efforts ahead of the 2022 midterms.
Limiting Misinformation on YouTube: After the 2018 launch of Change the Terms, the coalition regularly called on YouTube to strengthen its policies to rein in violent, hateful lies. In 2019, YouTube committed to stop recommending content that contained misinformation. It later announced that it would crack down on neo-Nazi content more aggressively.²⁴

Meaningful platform reforms require ongoing advocacy from external experts and activists, as well as scrutiny from independent researchers with comprehensive access to platform data. Therefore, it is essential to continually put pressure on social-media platforms and other tech companies to equitably and effectively protect the integrity of their products. Absent external accountability and inquiry, we would know even less about these companies’ opaque practices.²⁵

Free Press investigated the state of platform integrity at major tech companies in the 2022 report Empty Promises, in which we reviewed the policies of the four largest social-media platforms to consider how prepared, both in writing and in practice, the companies were for the 2022 midterm elections.²⁶ Our research found that although the largest tech companies long promised to fight disinformation and hate on their platforms, they failed to take adequate measures in the run-up to the 2022 midterms.

Companies like Meta, TikTok, Twitter and YouTube have failed to put sufficient policies, practices, automated systems and human capital in place to materially mitigate harm ahead of and during elections. To further complicate matters, these companies have created a labyrinth of commitments, announcements and policies that make it almost impossible to assess what they’re actually doing, if anything, to protect users.

After the 2022 midterms, the major platforms provided virtually no updates on the effectiveness of their policies nor any insights from their data about key network vulnerabilities. Meta provided no public summary. TikTok also failed to provide any publicly available summaries or reporting, though it issued brief community-guidelines updates. Twitter has no publicly available writing on the 2022 midterms. (Elon Musk took over Twitter just a week before these elections.) Alphabet, YouTube’s parent company, provided a short blog post about the 2022 election period, with no insights and scant details about YouTube.

In the absence of transparent reporting, one thing is certain: The platforms’ election-related policies and safety functions remain insufficient. ²⁷

Big tech's empty promises

$\"Big$

Taken together, these rollbacks over the past 12 months constitute an undeniable Big Tech backslide. This backsliding fosters less accountability across each of these platforms as companies turn their backs on years of evidence pointing to the crucial and outsized role they play in bringing people information, shifting their attitudes, and shaping discourse that affects civic engagement and democracy.

Of the platforms we examined, Twitter has rolled back the most policies and conducted the greatest ratio of layoffs-to-total-staff size. Meta has been a close second. And while YouTube had the fewest number of rollbacks over the last year, its policies were the weakest to begin with, as Free Press documented in earlier research. This points to the need for YouTube to reinstate and strengthen its policies overall.²⁸

These companies could easily reinstate their protective measures before 2024 — and it’s essential for them to do so, particularly in light of disinformation proliferating about the Israel-Hamas war. This report is an urgent invitation for Big Tech platforms to do more to protect users and democracies in 2024. We’re calling on the platforms to implement a specific set of recommendations before the U.S. presidential primary season begins in February 2024.

This chart documents the policies Meta, Twitter and YouTube have rolled back:

This backslide is not neutral in nature. The policies and teams that platforms deprioritized are key to understanding what these companies value and what they do not. All of the policy rollbacks across Meta, Twitter and YouTube deemphasize user safety and platform integrity, creating an opening for lies, hate and harassment to thrive. These three platforms have collectively laid off at least 40,750 workers, with massive cuts to trust and safety, content moderation, ethical AI and other teams tasked with maintaining user safety, content moderation and overall platform functionality.

Meta, Twitter and YouTube retreated from promises, policies and other actions to mitigate harm on their platforms in three main ways:

Policy removal: From Nov. 1, 2022 to Nov. 1, 2023, Meta, Twitter and YouTube removed a total of 17 policies they had had in place prior to the 2022 midterm elections.
Mass layoffs: All three companies have laid off tens of thousands of employees, totaling more than 40,750 across the three companies. Significant numbers of cuts were in the trust and safety, ethical engineering or responsible innovation, and content-moderation/consultant categories. Twitter removed its trust and safety team altogether.
Reinstating banned accounts: Meta, Twitter and YouTube reinstated Donald Trump’s accounts. Twitter also reinstated thousands of previously suspended accounts, including those belonging to white supremacists, conspiracy theorists, misogynists and others promoting hateful rhetoric.

$\"Icon$

$\"Material$

$\"\"/$

Understanding the Nature of Platform Rollbacks

With dozens of national elections happening around the world in 2024, platform-integrity commitments are more important than ever. However, major social-media companies are not remotely prepared for the upcoming election cycle.

Free Press has observed a notable drop in the promises these companies are making to users as well as a significant rollback in concrete measures companies once had in place. Meta, Twitter and YouTube have all removed long-standing and critical policies, laid off staff and entire teams, and reinstated and even monetized violative accounts.

Big Tech backslide

$\"\"/$

Reinstated Trump

$\"\"/$

Laid Off Content Moderators And/Or Trust & Safety Teams

new generative AI features on Meta platforms will draw on user data to train AI models.

$\"\"/$

Weakened Privacy & AI Policies

Stopped Moderating Covid Lies

$\"\"/$

stopped enforcing political ads policy, leaving an opening for bad actors to push lies in ads, which do not receive the same moderation treatment as user content

$\"\"/$

Weakened Political-Ad Policies

Meta Policy Rollbacks

$\"Twitter/X$

began allowing users to opt out of its fact-checking program

$\"\"/$

Imposed User Fact-Checking Limits

$\"\"/$

Reinstated Trump

$\"\"/$

Laid Off Content Moderators And/Or Trust & Safety Teams

new generative AI features on Meta platforms will draw on user data to train AI models.

$\"\"/$

Weakened Privacy & AI Policies

Stopped Moderating Covid Lies

$\"\"/$

stopped enforcing political ads policy, leaving an opening for bad actors to push lies in ads, which do not receive the same moderation treatment as user content

$\"\"/$

Weakened Political-Ad Policies

Twitter Policy Rollbacks

$\"Twitter/X$

began allowing users to opt out of its fact-checking program

$\"\"/$

Imposed User Fact-Checking Limits

Social-media companies’ written policies should provide clarity for users. These policies detail the way platforms moderate content, enforce their rules, deploy automated tools, and otherwise interact with users and their content to keep their services functional, authentic and useful. But these policies are inaccessible to most users. Instead of housing their rules in a centralized and accessible location, platforms often present them in an unruly patchwork of terms of service, community guidelines and standards, blog posts and tweets.

For well over a decade, civil-rights groups, civil-society organizations, lawmakers, tech ethicists and researchers have all tried to counsel the largest social-media platforms to make these policies clear, enforceable and equitable. They’ve urged platforms to protect public safety, public health, democracy and free expression. Change the Terms, a coalition anchored by Free Press, the Center for American Progress and the Global Project Against Hate and Extremism, was founded to disrupt platforms’ amplification of hate, extremism and lies, which plague the internet and endanger targeted groups.²⁹

Despite our best efforts, Meta, Twitter and YouTube removed a total of 17 policies between Nov. 1, 2022 and Nov. 1, 2023.

These actions signal a troubling step backward. Election lies, COVID and wartime disinformation, and harassment remain threats to public safety, public health and democracy. The major platforms were right to develop policies and procedures to flag, review, downrank and sometimes remove this content. Rolling back these policies creates a disaster each time a major current event captures public attention.³⁰

Policy Rollbacks Jeopardize Platform Integrity

$\"\"/$

Since November 2022, Alphabet, Meta and Twitter have collectively laid off at least 40,750 employees and contractors, prompting concern that these companies no longer have sufficient staff in place to effectively maintain platform health and safety.

In the first month of Musk’s ownership of Twitter, he gutted staff across some of the most critical teams that ensure a healthy and functional platform. Musk removed the board of directors and the Trust and Safety Council. He fired trust and safety staff, ethical engineering teams and content-moderation contractors. With fewer people on board to maintain Twitter’s integrity, many of the platform’s core capabilities buckled.³¹ Hate speech and disinformation spiked in the weeks and months that followed.³²

In March 2023, Alphabet started letting go significant numbers of employees on the ethics and safety teams at YouTube.³³ Meta announced its first round of job cuts the same month, with subsequent layoffs in April that had an “outsized effect on the company’s trust and safety work.”³⁴

Twitter has laid off approximately 7,000 people, or 82 percent of its staff. Alphabet, YouTube’s parent company, has laid off approximately 12,600 people. With almost no transparency about the implications of these Alphabet job losses for the YouTube platform, it’s unclear where the specific cuts took place, although the layoffs included some YouTube consultants. Meta has laid off approximately 21,000 people, or 25 percent of its workforce.³⁵

Where platforms choose to trim is a clear indication of company values. Meta has admitted that safety remains a “cost center…not a growth center.”³⁶ Fewer people often means less-effective moderation, leaving violative content on platforms longer. Platform-integrity failures during the first weeks of the Israel-Hamas war point to the critical role that staff in trust and safety and content moderation play during real-time crises.³⁷

Mass Layoffs Erode Platform Functionality & Moderation

$\"Material$

Is TikTok checking all the boxes?

Noticeably absent from our backsliders list is TikTok. TikTok is the only platform we analyzed that has not rolled back any major content-moderation policies since October 2022. And while it’s had some staff shifts, TikTok has not undertaken the same kinds of mass layoffs as its competitors.³⁸

During the last year, as competing platforms have rolled back content-moderation policies and protections for users, TikTok has expanded certain policies. For example, in March 2023, TikTok actually clarified and expanded existing moderation policies, and added new policies requiring disclosures around certain AI-generated content. The platform also introduced a new climate-misinformation policy.³⁹ Notably, TikTok made these changes just before CEO Shou Zi Chew appeared before Congress to testify about the company’s business and moderation practices.⁴⁰

Twitter’s November reinstatement of Trump prompted a domino effect across the platforms — on Jan. 25, Meta announced that it would reinstate Trump’s accounts.⁴⁷ On March 17, YouTube also reinstated Trump’s account.⁴⁸ Musk’s reinstatement provided the necessary cover for other platforms to do the same, even though Trump continues to spread lies about the 2020 election result and other election issues.⁴⁹

Conservative judge J. Michael Luttig testified before Congress that Trump and his base remain “a clear and present danger to American democracy” because of their potential to threaten the integrity of the 2024 election.⁵⁰

Musk’s sweeping amnesty reinstated thousands of previously banned accounts, including those belonging to white supremacists, conspiracy theorists and others who use the platform to sow division, spread lies, instigate violence and undermine democracy. A sample of reinstated accounts includes:

Andrew Anglin, an American neo-Nazi who founded the Daily Stormer, a website taking its name from the Nazi propaganda sheet known as Der Stürmer.⁵¹
Laura Loomer, a far-right political figure and self-proclaimed “proud Islamophobe” who unsuccessfully ran for a Florida congressional seat.⁵²
Andrew Tate, the influencer and former kickboxer known for posting extreme misogynistic videos. He has said that rape victims “bear some responsibility” for being raped and that he would threaten with a machete women who accuse him of cheating.⁵³
Anthime Gionet, known as Baked Alaska, a white-supremacist internet personality who attended the 2017 “Unite the Right” rally in Charlottesville. He has been sentenced for participating in the Jan. 6 insurrection and live-streaming the attack to his social-media followers.⁵⁴
Emerald Robinson, the former Newsmax reporter who has claimed that the COVID vaccine contains a satanic marker.⁵⁵
Gateway Pundit, an online media outlet notorious for promoting conspiracy theories related to vote tampering, climate change and COVID-19.

Meta also has a troubling history of cherry-picking VIP user accounts that amplify lies. For example, in August, Meta rejected its oversight board’s recommendation that it suspend the former Cambodian prime minister for video content that “included violent threats” toward his political opponents. The Hill reported that the former prime minister “had preemptively removed his Facebook page after the Oversight Board recommendation in June, and banished Facebook representatives from operating in the country.”⁵⁶

There are real free-expression concerns when it comes to how platforms limit and suspend accounts. Users may have an interest in the content of a particular suspended user. A political figure like Donald Trump may carry unique political and cultural interest for a large swath of the voting public. Platforms must balance these legitimate considerations against policies designed to minimize the spread of hateful speech, harassment, extremism, incitement to violence and lies.⁵⁷

As the platforms weigh reinstatements, the return of previously suspended accounts brings user eyeballs and money for the companies. But it also erodes user experience of these products, as the return of Trump and others ushers in more toxicity and lies.

$\"Rectangle\"/$

When the powerful use their digital platforms to promote hate, bigotry, lies and other vitriol, there’s often an outsized impact offline.⁴¹ For instance, researchers have established the link between Donald Trump’s online speech and offline violence in various contexts.⁴² One study found direct ties between Trump’s anti-Muslim tweets and a rise in anti-Muslim sentiment and hate crimes.⁴³ Furthermore, the House Select Committee investigated the role that social media played in the insurrection and determined that Trump’s posts on social media “set in motion a chain of events that led directly to the attack on the U.S. Capitol.”⁴⁴ Therefore, it is critical for platforms to maintain rules that apply equitably to all users and to enforce those policies irrespective of the influence certain users carry.

Despite this evidence, every major platform has reinstated Donald Trump over the last year. On Nov. 18, 2022, Elon Musk used a Twitter poll to ask his followers whether to reinstate Trump — a repeat offender of platform policies.⁴⁵ In other words, Musk outsourced a major policy decision to users and bots, pretending that he was allowing Twitter’s users to choose in the name of free speech. The poll received a slim margin in favor of reinstatement — likely boosted by fake accounts and bots⁴⁶ supporting Trump’s return.

Reinstatement & Monetization of Dangerous and Extremist Accounts

$\"Icon$

A League of Its Own: Twitter’s Total Failure

It’s a colossal understatement to say that Elon Musk has failed as the head of Twitter. Musk has destroyed almost everything that once made the platform worthwhile.

As soon as Musk took over Twitter, Free Press and many others expressed concern that his reckless decisions for the platform would harm people in the real world. There were warning signs from the outset: Use of the N-word surged immediately after Musk’s purchase last October, allowing bad actors to test the limits of the platform’s moderation systems.⁵⁸

Musk began by gutting content-moderation policies and decisions, ranging from ending the COVID-19 disinformation policy to discontinuing the trust and safety council.⁵⁹ Musk also laid off key staff on the public policy and the Machine Learning Ethics, Transparency, and Accountability teams, among others.⁶⁰ Just four weeks after Twitter’s purchase, Musk announced a “general amnesty,” swiftly reinstating thousands of previously banned accounts, including those belonging to prominent hate superspreaders.⁶¹

Over the subsequent months, Musk rolled back policies, staffing, design and other functions core to the platform’s ability to maintain healthy user feeds. Many of these rollbacks reversed years of progress Twitter had made in consultation with civil- and human-rights groups, including the Change the Terms coalition and organizations like Free Press.

Musk discontinued the deadnaming policy that it launched in 2018, which had previously banned users from calling a transgender person by an incorrect name, such as a no-longer-used birth name. He rolled back the platform’s misinformation policy and brought back political ads after former Twitter CEO Jack Dorsey sunsetted them for being a vehicle for political lies to enter user feeds unmoderated.

Musk’s rollbacks have led to a rise in toxicity across the platform. Six months into his ownership, researchers found that Twitter’s content-moderation rollbacks significantly increased hate speech across the platform compared to a similar timeframe prior to Musk’s purchase.⁶² Later results from a 2023 Trustlab study carried out for the European Commission reveal that Twitter carried the greatest amount of deceitful content when compared to other large social networks, including Facebook, Instagram, LinkedIn, TikTok and YouTube.⁶³ The study examined more than 6,000 unique social-media posts across platforms and found Twitter carried the most unmoderated disinformation.⁶⁴

In crisis moments, the platform’s failures have been catastrophic. During the fall of 2023, amid the ongoing Israel-Hamas war, the platform’s algorithms have boosted violent and disturbing images — some real, some faked — and disinformation about the conflict has spread across Twitter and migrated to mainstream news outlets. With few left at the company to vet questionable and violent content, posts are often left unchecked to spread like wildfire.

Furthermore, Musk’s decision to give special prominence to content by blue checkmark accounts without adequately verifying users’ identities — undoing years of effort to build trust on the platform — has given a soapbox to all sorts of grifters, conspiracy theorists and propagandists seeking to drown public discourse in lies coming from both sides of the conflict and many points in between.⁶⁵

Musk pushed the burden of fact checking onto Twitter users by encouraging them to use “Community Notes” to moderate the platform. But the sheer volume of fake reports following October’s Hamas attack in Israel stretched well beyond the reach of any user-powered fact checking. And it got so bad that former Twitter insiders who watchdog the feature told WIRED that Community Notes itself became a vehicle for spreading lies about the conflict. “A reliance on Community Notes is not good,” one of them said. “It’s not a replacement for proper content moderation.”⁶⁶ In crisis and wartime, the core functions of a platform matter more than ever. A platform must have well-trained moderators to enforce robust written policies, paired with strong user features to prevent network manipulation. Without any of these in place, it’s nearly impossible to distinguish fact from fiction on today’s Twitter.⁶⁷

Musk’s earliest content-moderation rollbacks sparked the #StopToxicTwitter pressure campaign, which has helped push more than half the platform’s advertisers to abandon the platform, with some leaving quietly and others citing concerns about their ads running alongside toxic content.⁶⁸ The cost of failure is not just financial. Communities, families, journalists and leaders have all suffered as a result of the platform’s failures in recent months.

But Threads launched just months after Meta announced a round of mass layoffs,⁷¹ as well as a hiring freeze, raising concerns about the company’s capacity to implement policies regarding the governance, moderation and security of a wholly new platform. According to Reuters,⁷² these layoffs directly impacted Meta’s privacy and integrity teams.

Free Press and two dozen civil-rights groups wrote to Meta⁷³ requesting details on how Threads would be moderated, how its content-moderation and privacy policies would be distinct from those of other Meta products, and how the platform planned to enforce policies with transparency. Meta’s boilerplate response — received over a month later than requested — failed to concretely answer any of our questions. In some instances, Meta simply pointed back to its own blog posts and other written documents, which lacked meaningful details.

Months after launch, Meta added a search function on Threads but has blocked various kinds of content. For example, it has blocked users from searching for content about COVID-19 and vaccines — which it deemed “potentially sensitive content.”⁷⁴

$\"Logos$

Threads Launch Spreads Meta Too Thin

Meta introduced Threads as “a new app, built by the Instagram team” — a team already responsible for managing a product with 2-billion monthly active users.⁶⁹ By directly linking this new platform to Instagram (users can’t delete their Threads accounts unless they’re willing to sacrifice their Instagram accounts, too), it’s clear that Meta hopes to capitalize on Instagram’s substantial global user base — a plan that, in the app’s early weeks, seemed successful.⁷⁰

AI to the Rescue?

Without staff on hand to fulfill core platform functions, executives may give outsized duties to artificial intelligence tools to manage mechanisms such as content moderation and review of flagged content. But even with the assistance of AI, tech companies are unable to enforce content-moderation policies at scale. Mass layoffs only exacerbated this problem.

Furthermore, AI and other automated tools these companies use simply don’t have adequate cultural nuance to do it all. Proper and extensive auditing of automated tools requires humans to review the results. These companies cannot expect automated tools — absent the adequate staffing to train and review AI processes — to effectively maintain platform integrity.

The potential for platform misuse of AI technology should raise several flags for lawmakers and regulators. Platforms that use any digital automation often train their algorithms using untold amounts of user data. This data can include users’ names, addresses, purchasing histories, financial information and other sensitive information such as Social Security numbers, medical records — and even people’s biometric data, like fingerprints and iris recognition.

Companies like Meta that deal in sophisticated algorithms and AI often say they’re gathering this information to deliver hyper-personalized and “improved” experiences for people. But dangerous consequences flow from having advanced algorithms analyze our data without the proper guardrails and auditing of these tools. These companies can also use this data processing to exclude specific users from receiving critical election information, and they can target users to receive disinformation about voting locations and candidates. Simply put, unchecked AI violates our digital civil rights.

Tech companies and lawmakers have different — though equally necessary — roles to properly rein in abusive and discriminatory AI tools.⁷⁵ Private companies, particularly social-media companies, must audit and review the AI tools they employ, with adequate human review of the impact of automated processes. Private companies should gather the minimum data about users, with stringent data-conservative approaches to the use and collection of that data. Lawmakers and regulators should also mandate transparency from private companies, create data-minimization requirements and eliminate algorithmic discrimination.

1 cross-platform reinstatement

hundreds of reinstatements on Twitter

POLICY

ROLLBACKS

PROPOSED POLICY ROLLBACKS

LAYOFFS

PR STUNT

$\"Icon$

LAYOFFS

$\"Icon$

POLICY ROLLBACK

$\"Icon$

REINSTATEMENT

$\"Icon$

LITIGATION

$\"Icon$

Musk meets with civil-rights leaders, makes empty promises

November 2, 2022

[endnote 76]

Musk cuts 50% of Twitter’s workforce, including many trust and safety, ethical AI, marketing, and public policy employees

November 4, 2022

[endnote 77]

Meta lays off 13 percent of workforce

November 9, 2022

[endnote 78]

Musk launches paid-subscription program Twitter Blue, ending use of verified blue checkmarks

November 9, 2022

[endnote 79]

Musk announces that Twitter will reinstate Donald Trump’s account

November 19, 2022

[endnote 80]

Musk announces “general amnesty” for previously banned accounts from neo-Nazis and other extremists

November 24, 2022

[endnote 81]

Musk eliminates Twitter’s ban on COVID-19 disinformation

November 29, 2022

[endnote 82]

Musk eliminates Twitter’s Trust and Safety Council

December 12, 2022

[endnote 83]

Alphabet, which owns YouTube, lays off 12,000 employees

January 20, 2023

[endnote 84]

Meta announces that it’s reinstating Donald Trump’s account

January 25, 2022

[endnote 85]

Google cuts a third of Jigsaw staff who prioritized fighting disinformation and online toxicity

February 3, 2023

[endnote 86]

Twitter lays off product, data science, engineering and site reliability workers

February 24, 2023

[endnote 87]

Google cuts staff on ethical AI and trust and safety teams

March 13, 2023

[endnote 88]

YouTube reinstates Donald Trump’s accounts

March 17, 2023

[endnote 89]

Twitter guts deadnaming policy, allowing deadnaming and misgendering of trans users

April 8, 2023

[endnote 90]

Twitter updates enforcement philosophy to “freedom of speech, not reach”

April 17, 2023

[endnote 91]

Meta conducts more layoffs, with outsized impact on trust and safety

April 19, 2023

[endnote 92]

TikTok’s U.S. head of trust and safety leaves the company

May 2, 2023

[endnote 93]

Business Insider reports that Musk has laid off all but 1,000 employees

May 2, 2023

[endnote 94]

Google cuts contractors who had worked on YouTube services

May 15, 2023

[endnote 95]

Meta lays off about 6,000 people, totaling roughly 21,000 layoffs since November 2022

May 24, 2023

[endnote 96]

YouTube stops removing Big Lie content

June 2, 2023

[endnote 97]

Meta rolls back COVID-19 content-moderation policies

June 16, 2023

[endnote 98]

Meta launches Threads

July 5, 2023

[endnote 99]

Musk rebrands Twitter as X

July 24, 2023

[endnote 100]

Musk announces plan to remove block feature on Twitter

August 18, 2023

[endnote 101]

Meta announces that it will allow people to opt out of fact-checking program

August 25, 2023

[endnote 102]

Twitter decides to allow political ads on the platform

August 29, 2023

[endnote 103]

YouTube weakens strike policy for violative video content

August 29, 2023

[endnote 104]

Twitter guts privacy protections and will use all user data — including DMs and biometrics — to train its AI model

August 31, 2023

[endnote 105]

WIRED reports that Meta is not enforcing its political-ads policy

September 1, 2023

[endnote 106]

Musk considers filing a defamation lawsuit against ADL

September 3, 2023

[endnote 107]

Musk sues California over its transparency law

September 8, 2023

[endnote 108]

Threads blocks searches for COVID, vaccines and related information

September 11, 2023

[endnote 109]

Google lays off hundreds of recruiters

September 13, 2023

[endnote 110]

Musk proposes charging all Twitter users to access the platform

September 18, 2023

[endnote 111]

Google opens access for AI tool Bard to user data on YouTube

September 19, 2023

[endnote 112]

Twitter globally disables feature for reporting election disinformation except in the EU per statutory requirement there

September 26, 2023

[endnote 113]

YouTube relaxes advertising policy to allow more monetization without penalty

September 26, 2023

[endnote 114]

Musk cuts half of Twitter’s global election-integrity teams

September 27, 2023

[endnote 115]

Meta’s new generative AI features will draw on user data to train AI models

September 27, 2023

[endnote 116]

Musk removes the headline feature for links shared

October 4, 2023

[endnote 117]

Meta cuts staff from the Facebook Agile Silicon Team, tasked with virtual-reality functions

October 4, 2023

[endnote 118]

A Year of Backsliding: Timeline

Cross-Sector Attacks on Tech Accountability

To make matters worse, Republicans in Congress have issued similar threats — as well as subpoenas — to stifle research about the spread of election disinformation.¹²⁶

This past summer, Republican House Judiciary Committee Chairman Jim Jordan led an effort demanding documents from — and meetings with — leading U.S. academic researchers who examine disinformation. Jordan and his allies have accused these researchers of “colluding with government officials to suppress conservative speech.”¹²⁷ These attacks have led researchers to retreat from more public advocacy about the need for tech accountability, with some citing the attacks as the impetus to step back.¹²⁸

As we approach the 2024 election season, we need more research and transparency around what the platforms are doing. Cordoning off communication between the tech industry and other sectors — including government, civil society and researchers — will likely give the platforms more room to back away from their previous promises and policies. We should find ways to better coordinate with social-media companies to increase the integrity of information on their platforms. In our Recommendations section, we lay out the essential steps tech companies must take in the coming months.

Strategic Lawsuits Against Platform Transparency

Most troubling is a set of lawsuits Musk has initiated to silence researchers and critics.

So far, he has brought two lawsuits, one challenging the Center for Countering Digital Hate (CCDH) and another challenging a new transparency statute in California.¹²⁰ He has threatened to bring a third suit against the Anti-Defamation League (ADL).

In July, Musk sued CCDH, alleging the nonprofit committed computer fraud in its use of a tool that monitors advertisements on Twitter.¹²¹ CCDH has refuted Musk’s claims and refused to buckle under the pressure of the lawsuit, and allies have sounded the alarm that this suit is a dangerous attempt to chill the organization’s independent research.

Musk later threatened to sue the ADL for supporting and encouraging companies to pull their ad spending from Twitter. Musk withdrew his threat once the ADL denied its involvement in such efforts and affirmed that it would continue advertising on the platform. Musk’s success in getting the ADL to cave to his demands will likely lead to more threats to researchers and the tech-accountability field more broadly.

Musk has also sued California over a new law mandating semiannual reports from major social-media platforms that describe their content-moderation practices and share data on the numbers of objectionable posts and how companies addressed them.¹²² The law also requires companies to make public their terms of service. Failure to comply risks civil fines of up to $15,000 a day for each violation. Musk and Twitter claim that the law violates the platform’s First Amendment rights. A federal court in Sacramento is currently assessing this argument.¹²³

The use of litigation to silence critics is an old tactic.¹²⁴ Trump regularly uses this strategy: Media-law specialist Susan Seager notes that “Trump and his companies have been involved in a mind-boggling 4,000 lawsuits over the last 30 years and sent countless threatening cease-and-desist letters to journalists and critics.”¹²⁵ Musk’s use of this Trumpian tactic is hardly surprising — the two men share numerous hallmarks of authoritarian bullies, including claiming to champion free speech while silencing speech they dislike.

Strategic lawsuits attacking platform transparency are similar to “SLAPP” suits (strategic lawsuits against public participation), which chill public participation and journalistic inquiry. The use of these lawsuits is dangerous for researchers who might otherwise want to investigate platform behavior but decide not to out of fear of being sued. It’s also dangerous for the public, which might remain in the dark about tech companies’ unethical practices. We need urgent data and information-sharing from platforms and thoughtful collaboration across sectors to minimize threats to national security, democratic institutions and local communities that originate on platforms.

Frances Haugen testified before the U.S. Senate on Oct. 5, 2021. Original photo by U.S. Senate Committee on Commerce via Wikimedia Commons

$\"Frances$

transparency and equity. Absent much-needed regulatory oversight, these companies are doing less and less to maintain platform integrity.

Free Press has also documented a different but equally grave set of threats to tech-accountability work at organizations like ours.

The rollbacks described above — symptoms of a broader backslide at the largest social-media companies — are but one piece of a growing tech-accountability problem.

Two years ago, Frances Haugen testified before Congress, where she offered bombshell evidence on when and how much Meta executives knew about the extent to which their platforms were undermining democracy and public safety. Facebook, now Meta, chose profits over user safety over and over, making moderation and design decisions that kept people engaged with toxic content — and boosted the company’s bottom line.¹¹⁹

The rollbacks described above — symptoms of a broader backslide at the largest social-media companies — are but one piece of a growing tech-accountability problem.

These findings prompted a number of congressional hearings, legislative proposals, academic research reports and civil-society recommendations — all undergirding years of advocacy from those seeking platform accountability, transparency and equity. Absent much-needed regulatory oversight, these companies are doing less and less to maintain platform integrity.

Free Press has also documented a different but equally grave set of threats to tech-accountability work at organizations like ours.

Frances Haugen testified before the U.S. Senate on Oct. 5, 2021. Original photo by U.S. Senate Committee on Commerce via Wikimedia Commons

$\"Frances$

Frances Haugen testified before the U.S. Senate on Oct. 5, 2021. Original photo by U.S. Senate Committee on Commerce via Wikimedia Commons

$\"Frances$

The rollbacks described above — symptoms of a broader backslide at the largest social-media companies — are but one piece of a growing tech-accountability problem.

Free Press has also documented a different but equally grave set of threats to tech-accountability work at organizations like ours.

The forecast ahead

What Governments MUST Do

The platforms alone bear responsibility for content moderation — and Free Press Action opposes government efforts to dictate those content standards. But there are steps governments should take to prevent fraud and scams — and to protect democracy, public safety, and human and civil rights online.

Thanks to strict regulatory requirements in places like the European Union, the cost of social-media companies doing less is steep.¹³⁷ Here in the United States, we desperately need meaningful regulation to rein in social-media platforms’ destructive and reckless behavior. Free Press Action is calling on U.S. lawmakers and regulators to codify reforms that:

Minimize data that companies collect and retain to protect against discriminatory targeting of users with tailored content and advertising;
Ban algorithmic discrimination by platforms and other internet services that use AI tools to target users;
Require regular platform transparency and disclosure reports on content-virality trends, results of AI decision-making tools, and visibility and take-downs of political ads — all across languages;
Develop a private civil right of action for violations that flow from platforms’ use of personal sensitive data on users; and
Leverage agency and White House authority to pursue accountability at the Federal Trade Commission, Department of Justice, Federal Election Commission and other relevant agencies to craft new rules and launch investigations and prosecutions where statutory violations arise.

As Musk has sunk Twitter down a black hole, his bottom line has suffered to the point where the platform is not worth even a fraction of the $44 billion he paid to purchase it.¹³⁸ The cost of doing less is in the billions. But if the last several years have taught us anything, it’s that content moderation isn’t just about social-media companies’ bottom lines.

What Platforms Must Do

As with every election cycle, threats abound. Civil-society groups have sounded the alarm about what kinds of threats we might see in 2024:¹²⁹

Supercharged disinformation and fake news stories spun up faster than ever before due to efficient artificial intelligence tools that can personalize disinformation, amplify calls for violence and dissuade civic engagement
AI-generated imagery, audio and videos, known as deepfakes, which mislead voters about candidate comments, positions and other attitudinal markers to sway their voting preferences
Laser-targeted political advertising zeroing in on protected classes in discriminatory and exploitative ways
Tailor-made content crafted to discourage certain categories of voters from participating, based on data platforms have already collected about them
Incorrect labeling of imagery and/or AI-generated false imagery about specific polling locations

Social-media companies tend to treat the threats posed around election cycles as anecdotal and time-limited. Free Press and the broader civil- and human-rights field have urged platforms to consider the year-round vulnerabilities that allow election-specific content to prey on voters. Platforms’ governance and enforcement decisions cannot adequately guard against manipulations if spun up a month or two before an election and taken down immediately after the polls close.

We urge social-media platforms to take the following steps to protect users on their platforms:

Reinvest in staffing and teams needed to safeguard election integrity, trust and safety, and moderation. The major platforms have laid off more than 40,750 employees in the last year alone. At least several thousand of those were positions critical to the moderation and enforcement of platforms’ policies — policies that are essential to keeping lies and extremist rhetoric in check. Without the staff needed to adequately confirm content-moderation decisions triggered by automated review, dangerous online discourse will likely seep into more users’ feeds — even when deemed violative — for longer periods. This will happen because these platforms simply don’t have the staff to efficiently and effectively moderate content.
Reinstate disinformation policies — including those governing election and COVID disinformation — and bolster policy moderation to limit exposure to violence and lies. Platforms typically pull together rapid-response moderation and other product features once a crisis unfolds, which is too late. These companies typically roll out election-integrity efforts mere months before an election. This is also too late. And this negligence has real-world consequences: The Global Project Against Hate and Extremism, a nonprofit research organization, found that only 41 percent of voters feel safe or very safe at their polling place, and only a quarter of young people, 28 percent of Black people and 37 percent of Latinx people feel safe or very safe at their polling places.¹³⁰ Researchers at UT Austin have noted that racialized disinformation and intimidation tactics target Black, Latinx and Indigenous voters ahead of election cycles.¹³¹ As these kinds of threats, dangerous rhetoric and lies reach people online, they can have an impact on how people behave — and whether they vote — on Election Day. If left unchecked, misleading and violent content may discourage voters from participating in the democratic process.
Launch 2024 election-specific platform interventions in time for the U.S. primaries and keep them in place through at least February 2025. This includes user-information portals for voters to get real-time information about voter registration, polling locations, ways to access credible civic-engagement tools and more. In previous election cycles, platforms have rolled out election-specific policy updates and announcements the summer before a fall election. This is simply too late. These announcements — and the companion interventions needed to safeguard user safety and civic engagement — must launch by February 2024. These interventions must stay in place long after Election Day to protect the election results and safeguard against attempts to undermine them.
Hold VIP accounts to the same standards as those of other users. Companies should hold candidates, celebrities and other public figures to account when they break company rules. Meta’s own Facebook Oversight Board has documented unequal treatment of users, with some users subject to more lenient review policies while “layperson” accounts are more stringently moderated according to Meta’s cross-check program.¹³² Free Press has previously documented this special-treatment phenomenon.¹³³ It persists today, with platforms giving user accounts like Donald Trump’s great leniency to promote false claims that the 2020 election was stolen.¹³⁴ Companies need to hold these users to the same — or even stricter — moderation and enforcement review standards as their layperson counterparts.
Develop more efficient review and enforcement on political-ad content across languages. No major social-media platform has a streamlined database that allows one to identify and analyze political ads’ visibility, veracity, spending and more.¹³⁵ More efficient human review of political ads across languages must occur prior to AI analysis and review. Companies must also ensure human enforcement of their mis- and disinformation and extremism policies on those ads, with timely and transparent data shared to external sectors about trends.
Develop better transparency and disclosure policies and regularly share core metrics data with researchers, journalists, lawmakers and the public. These companies should take action, as promised in their terms of service, on violative content and on tracking core metrics to distribute externally.¹³⁶ They should provide affordable and comprehensive API access to researchers and others. They should share audit reports of moderation and enforcement trends, as well as reporting on the impact of their automated tools.

$\"\"/$

Social-media companies’ retreat from accountability is happening against the backdrop of the 2024 elections. Dozens of national elections will occur across the world — and many people will get information about candidates, issues, voting logistics and election results on social-media platforms.

Recommendations for 2024

Vacation Days

Dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempordern for the next morum generation next.

Pet Lovers

Dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempordern for the next morum generation next.

To develop this research report, Free Press analyzed announcements from Alphabet, Meta, TikTok and Twitter as well as external media coverage of changes to corporate policies, layoffs, reinstatements and other measures the companies took between Nov. 1, 2022 and Nov. 1, 2023. The platforms’ lack of transparency necessitated this external research. For example, we tried to avoid conflicting reports without details about layoffs or dates. We also omitted mention of policy changes or other efforts that companies announced and then quickly reversed, because doing so would have skewed the analysis to focus disproportionately on Twitter. Moreover, including such policies would not provide an accurate landscape of a given platform’s policies current to the time of publication.

The report was written by Nora Benavidez, with research support and editing from Rose Lang-Maso. The development and production of this report could not have happened without support from Free Press colleagues Craig Aaron, Jessica J. González, Timothy Karr, Jenna Ruddock and Matt Wood. This report was edited by Amy Kroin and designed by Courtney Morrison, Sara Pritt and Dutch Cosmian.

Methodology & Acknowledgements

Endnotes:

Free Press, Empy Promises: Inside Big Tech’s Weak Effort to Fight Hate and Lies in 2022, Oct. 27, 2022: https://www.freepress.net/sites/default/files/2022-10/empty_promises_inside_big_techs_weak_effort_to_fight_hate_and_lies_in_2022_free_press_final.pdf, p. 8
Although Elon Musk rebranded Twitter as X in July 2023, it is still commonly referred to by its original name, including throughout this report.
Vittoria Elliott, “Meta Isn’t Enforcing Its Own Political Ads Policy, While the 2024 US Election Looms,” WIRED, Sept. 1, 2023, https://www.wired.com/story/meta-prageru-advertising/
Meta, “Privacy Matters: Meta’s Generative AI Features,” Sept. 27, 2023, https://about.fb.com/news/2023/09/privacy-matters-metas-generative-ai-features/; Sarah Perez, “X’s Privacy Policy Confirms It Will Use Public Data to Train AI Models,” TechCrunch, Sept. 1, 2023, https://techcrunch.com/2023/09/01/xs-privacy-policy-confirms-it-will-use-public-data-to-train-ai-models/; Michael Leidtke, “Google Brings Its AI Chatbot Bard Into Its Inner Circle, Opening Door to Gmail, Maps, YouTube,” Associated Press, Sept. 19, 2023, https://apnews.com/article/google-artificial-intelligence-bard-gmail-youtube-maps-1229638b82d19afb5226c913821fa1ad
Dean Jackson, Meaghan Conry and Alex Newhouse, “Insiders’ View of the January 6th Committee’s Social Media Investigation,” Just Security, Jan. 5, 2023, https://www.justsecurity.org/84658/insiders-view-of-the-january-6th-committees-social-media-investigation/; Cat Zakrzewski, Cristiano Lima and Drew Harwell, “What the Jan. 6 Probe Found Out About Social Media, but Didn’t Report,” The Washington Post, Jan. 17, 2023, https://www.washingtonpost.com/technology/2023/01/17/jan6-committee-report-social-media/
Olivia Little, “TikTok Is Blocking Searches for WGA Amid Ongoing Writers’ Strike,” Media Matters for America, Sept. 11, 2023, https://www.mediamatters.org/tiktok/tiktok-blocking-searches-wga-amid-ongoing-writers-strike
Google laid off contractors from Accenture and Cognizant who had previously worked on platform services for YouTube. See Thomas Maxwell and Hugh Langley, “Google Is Downsizing Its Contract Workforce That Supports YouTube Shortly After One Contractor Team's Union Victory,” Business Insider, May 15, 2023, https://www.businessinsider.com/google-downsizes-contract-workforce-youtube-union-win-2023-5
Vittoria Elliott, “How X Is Suing Its Way Out of Accountability,” WIRED, Aug. 15, 2023, https://www.wired.com/story/twitter-x-ccdh-lawsuit-data-crackdown/
The European Union has examined Twitter’s failed content moderation around the Israel-Hamas war and is considering legal action for the platform hosting terrorist-organization content without proper moderation in place to remove it.
Social Media & the January 6th Attack on the U.S. Capitol: Summary of Investigative Findings, draft report, Tech Policy Press, https://techpolicy.press/wp-content/uploads/2023/01/J6-Committee-Draft-Social-Media-Report-TPP.pdf
“The bipartisan statement was signed by the Hoover Presidential Foundation, the Roosevelt Institute, the Truman Library Institute, the John F. Kennedy Library Foundation, the LBJ Foundation, the Richard Nixon Foundation, the Gerald R. Ford Presidential Foundation, the Carter Center, the Ronald Reagan Presidential Foundation and Institute, the George & Barbara Bush Foundation, the Clinton Foundation, the George W. Bush Presidential Center and the Obama Presidential Center.” See Gary Fields, “Presidential Centers from Hoover to Bush to Obama Unite to Warn of Fragile State of US Democracy,” Associated Press, Sept. 7, 2023, https://apnews.com/article/united-states-democracy-presidents-threats-joint-statement-5530a89df2c41d58a22961f63fb0e6ff
Global Project Against Hate and Extremism, “Violent Antisemitic and Anti-Muslim Hate Escalating Online in Wake of Hamas Attacks on Israel,” Oct. 12, 2023, https://globalextremism.org/post/violent-antisemitic-and-anti-muslim-hate-escalating-online-in-wake-of-hamas-attacks-on-israel/
Alice Speri, “‘Beheaded Babies’ Report Spread Wide and Fast — But Israel Military Won’t Confirm it,” The Intercept, Oct. 11, 2023, https://theintercept.com/2023/10/11/israel-hamas-disinformation/
Oliver Darcy, “Déjà Coup: How Election Lies Sparked the Violent Attack on Brazil’s Government,” CNN, Jan. 19, 2023, https://www.cnn.com/2023/01/09/media/brazil-government-reliable-sources/index.html
Enda Curran and Alan Crawford, “Brace for Elections: 40 Countries Are Voting in 2024,” Bloomberg, Nov. 1, 2023, https://www.bloomberg.com/news/articles/2023-11-01/brace-for-elections-40-countries-are-voting-in-2024
Marwa Fatawa, “Facebook Is Bad at Moderating in English. In Arabic, It’s a Disaster,” Rest Of World, Nov. 18, 2021, https://restofworld.org/2021/facebook-is-bad-at-moderating-in-english-in-arabic-its-a-disaster; Olivia Solon, “'Facebook Doesn't Care': Activists Say Accounts Removed Despite Zuckerberg's Free-Speech Stance,” NBC News, June 15, 2020, https://www.nbcnews.com/tech/tech-news/facebook-doesn-t-care-activists-say-accounts-removed-despite-zuckerberg-n1231110; Jacqueline Rowe, “Marginalised Languages and the Content Moderation Challenge,” Global Partners Digital, March 2, 2022, https://www.gp-digital.org/marginalised-languages-and-the-content-moderation-challenge
Salvador Rodriguez, “Senators Demand Facebook CEO Mark Zuckerberg Answer Questions After Whistleblower’s Revelations at Hearing,” CNBC, Oct. 5, 2021, https://www.cnbc.com/2021/10/05/congress-demands-mark-zuckerberg-answer-questions-at-haugen-hearing.html
“Luján, Klobuchar, Cárdenas Lead Colleagues Urging Tech CEOs to Combat Spanish-Language Disinformation,” July 30, 2021, https://www.lujan.senate.gov/newsroom/press-releases/lujan-klobuchar-cardenas-lead-colleagues-urging-tech-ceos-to-combat-spanish-language-disinformation/
Naomi Nix, “Inside the Civil Rights Campaign to Get Big Tech to Fight the Big Lie,” The Washington Post, Sept. 22, 2022, https://www.washingtonpost.com/technology/2022/09/22/midterms-elections-social-media-civil-rights/; Id, “Big Tech is Failing to Fight the Big Lie, Civil Rights Groups Charge,” The Washington Post, Oct. 27, 2022, https://www.washingtonpost.com/technology/2022/10/27/civil-rights-2022-midterms/
Joseph Cox and Jason Koebler, “Facebook Bans White Nationalism and White Separatism,” Motherboard, March 27, 2019, https://www.vice.com/en/article/nexpbx/facebook-bans-white-nationalism-and-white-separatism; “An Update to How We Address Movements and Organizations Tied to Violence,” Meta, Aug. 19, 2020, https://about.fb.com/news/2020/08/addressing-movements-and-organizations-tied-to-violence/
Adi Robertson, “Twitter Has Banned Misgendering or ‘Deadnaming’ Transgender People,” The Verge, Nov. 27, 2018, https://www.theverge.com/2018/11/27/18113344/twitter-trans-user-hateful-content-misgendering-deadnaming-ban
“Updating Our Rules Against Hateful Conduct,” Twitter, July 9, 2019, https://blog.twitter.com/en_us/topics/company/2019/hatefulconductupdate; Queenie Wong, Corinne Reichert and Oscar Gonzalez, “Twitter's Decision to Halt Political Ads Puts More Pressure on Facebook,” CNET, Oct. 30, 2019, https://www.cnet.com/tech/mobile/twitters-decision-to-halt-political-ads-puts-more-pressure-on-facebook/
“Updating Our Approach to Misleading Information,” Twitter, May 11, 2020, https://blog.twitter.com/en_us/topics/product/2020/updating-our-approach-to-misleading-information
Casey Newton, “YouTube Says It Will Recommend Fewer Videos About Conspiracy Theories,” The Verge, Jan. 25, 2019, https://www.theverge.com/2019/1/25/18197301/youtube-algorithm-conspiracy-theories-misinformation; Kevin Roose and Kate Conger, “YouTube to Remove Thousands of Videos Pushing Extreme Views,” The New York Times, June 5, 2019, https://www.nytimes.com/2019/06/05/business/youtube-remove-extremist-videos.html
Natalie Mathes, “Social Media Platforms Are Backsliding, Paving the Way for Chaos in 2024,” Media Matters for America, June 13, 2023, https://www.mediamatters.org/facebook/social-media-platforms-are-backsliding-paving-way-chaos-2024-; GLAAD, “Social Media Safety Index 2023,” https://glaad.org/publications/social-media-safety-index-2023/; Megan Shahi, “Protecting Democracy Online in 2024 and Beyond,” Center for American Progress, Sept. 14, 2023, https://www.americanprogress.org/article/protecting-democracy-online-in-2024-and-beyond/; Center for Countering Digital Hate, X Content Moderation Failure, Sept. 13, 2023, https://counterhate.com/research/twitter-x-continues-to-host-posts-reported-for-extreme-hate-speech/
Free Press, Empty Promises: Inside Big Tech’s Weak Effort to Fight Hate and Lies in 2022, Oct. 27, 2022, https://www.freepress.net/policy-library/empty-promises-inside-big-techs-weak-effort-fight-hate-and-lies-2022
Google, YouTube’s parent company, provided scant details on its 2022 election period; see Laurie Richardson, “Recapping Our Work on the 2022 U.S. Midterm Elections,” Google, https://blog.google/outreach-initiatives/civics/recapping-our-work-on-the-2022-us-midterm-elections/. Meta has no public summary of the 2022 midterm-election period. TikTok also failed to provide any publicly available summaries or reporting on the 2022 midterms, though it had brief community-guidelines updates; see “Community Guidelines Enforcement Report,” TikTok, https://www.tiktok.com/transparency/en-us/community-guidelines-enforcement-2022-4/. Twitter has no publicly available writing on the 2022 midterms. Elon Musk took over Twitter just a few days before the 2022 midterms and quickly removed critical staff on trust and safety and other teams.
Free Press, Empty Promises: Inside Big Tech’s Weak Effort to Fight Hate and Lies in 2022, Oct. 27, 2022, https://www.freepress.net/policy-library/empty-promises-inside-big-techs-weak-effort-fight-hate-and-lies-2022
Change the Terms, https://www.changetheterms.org/
Nora Benavidez, “Social Media Platform Integrity Matters in Times of War,” Just Security, Oct. 13, 2023, https://www.justsecurity.org/89449/social-media-platform-integrity-matters-in-times-of-war-now-more-than-ever/
Jennifer Korn, “As Twitter Failures Go from Bad to Worse, Users Wonder How Long It Can Stay Online,” CNN, March 12, 2023, https://www.cnn.com/2023/03/12/tech/twitter-breaking/index.html
Rashawn Ray and Joy Anyanwu, “Why Is Elon Musk’s Twitter Takeover Increasing Hate Speech?” Brookings Institution, Nov. 23, 2022, https://www.brookings.edu/articles/why-is-elon-musks-twitter-takeover-increasing-hate-speech/; Julia Cohen, “New Twitter, Now With More Hate,” USC Viterbi School of Engineering, April 20, 2023, https://viterbischool.usc.edu/news/2023/04/new-twitter-now-with-more-hate/; TrustLab, “Code of Practice on Disinformation,” Sept. 22, 2023, https://disinfocode.eu/wp-content/uploads/2023/09/code-of-practice-on-disinformation-september-22-2023.pdf
Hayden Field and Jonathan Vanian, “Tech Layoffs Ravage the Teams That Fight Online Misinformation and Hate Speech,” CNBC, May 26, 2023, https://www.cnbc.com/2023/05/26/tech-companies-are-laying-off-their-ethics-and-safety-teams-.html#:~:text=Twitter%20effectively%20disbanded%20its%20ethical,safety%20department%2C%20according%20to%20reports; Thomas Maxwell and Hugh Langley, “Google Is Downsizing Its Contract Workforce That Supports YouTube Shortly After One Contractor Team’s Union Victory,” Business Insider, May 15, 2023, https://www.businessinsider.com/google-downsizes-contract-workforce-youtube-union-win-2023-5
Hayden Field and Jonathan Vanian, “Tech Layoffs Ravage the Teams That Fight Online Misinformation and Hate Speech,” CNBC, May 26, 2023, https://www.cnbc.com/2023/05/26/tech-companies-are-laying-off-their-ethics-and-safety-teams-.html#:~:text=Twitter%20effectively%20disbanded%20its%20ethical,safety%20department%2C%20according%20to%20reports
Ryan Mac, Mike Isaac and Kate Conger, “Twitter Outages Are on the Rise Amid Elon Musk’s Job Cuts,” The New York Times, Feb. 8, 2023, https://www.nytimes.com/2023/02/28/technology/twitter-outages-elon-musk.html; Clare Duffy, “Meta’s Latest Round of Layoffs Is Underway,” CNN, April 19, 2023, https://www.cnn.com/2023/04/19/tech/meta-tech-team-layoffs-begin/index.html
Clare Duffy, Aditi Sangal, Melissa Mahtani and Meg Wagner, “Internal Facebook Documents Revealed,” CNN, Oct. 26, 2021, https://www.cnn.com/business/live-news/facebook-papers-internal-documents-10-25-21/h_c0b027ed77300bdfec6771e8c0788cff#:~:text=Former%20Facebook%20employee%20and%20whistleblower,very%20short%2Dterm%20in%20thinking
Pranav Dixit, “Social Media Platforms Swamped with Fake News on the Israel-Hamas War,” Al Jazeera, Oct. 10, 2023, https://www.aljazeera.com/news/2023/10/10/social-media-platforms-swamped-with-fake-news-on-the-israel-hamas-war; Nora Benavidez, “Social Media Platform Integrity Matters in Times of War,” Just Security, Oct. 13, 2023, https://www.justsecurity.org/89449/social-media-platform-integrity-matters-in-times-of-war-now-more-than-ever/
Alyssa Stringer, “A Comprehensive List of 2023 Tech Layoffs,” TechCrunch, Nov. 14, 2023, https://techcrunch.com/2023/11/14/tech-layoffs-2023-list/
Aisha Malik, “TikTok Overhauls Its Community Guidelines, Adds New Policies on AI and Climate Misinformation,” TechCrunch, March 21, 2023, https://techcrunch.com/2023/03/21/tiktok-overhauls-community-guidelines-adds-new-policies-ai-climate-misinformation/
Id.
Heidi Beirich and Wendy Via, Democracies Under Threat, Global Project Against Hate and Extremism, March 2021, https://globalextremism.org/wp-content/uploads/2022/01/GPAHE_Democracy-under-threat.pdf
Id.; Justin Hendrix, “The Final Jan. 6th Report on the Role of Social Media,” Tech Policy Press, Dec. 23, 2022, https://techpolicy.press/the-final-january-6th-report-on-the-role-of-social-media/
Karsten Müller and Carlo Schwarz, “From Hashtag to Hate Crime: Twitter and Antiminority Sentiment,” American Economic Journal: Applied Economics, Vol. 15, No. 3, July 2023, https://www.aeaweb.org/articles
Select Committee to Investigate the January 6th Attack on the United States Capitol, Final Report, Dec. 22, 2022, https://www.govinfo.gov/content/pkg/GPO-J6-REPORT/pdf/GPO-J6-REPORT.pdf; see pp. 55, 499
Elon Musk tweet, Twitter, Nov. 19, 2022, https://twitter.com/elonmusk/status/1594131768298315777
Davey Alba and Bloomberg, “The People Have Spoken—or Maybe Not: Elon Musk’s Use of Twitter Polls for Key Decisions Invites Manipulation,” Fortune, Dec. 22, 2022, https://fortune.com/2022/12/22/elon-musk-twitter-poll-manipulation-bots-acquisition/
Nick Clegg, “Ending Suspension of Trump’s Accounts With New Guardrails to Deter Repeat Offenses,” Meta, Jan. 25, 2023, https://about.fb.com/news/2023/01/trump-facebook-instagram-account-suspension/
BBC, “YouTube Reinstates Donald Trump's Channel,” March 17, 2023, https://www.bbc.com/news/technology-64993603
Martin Pengelly, “Trump a ‘Clear and Present Danger to US Democracy’, Conservative Judge Warns,” The Guardian, June 16, 2022, https://www.theguardian.com/us-news/2022/jun/16/trump-clear-present-danger-to-us-democracy-conservative-judge-warns
Id.
Nikki McCann Ramirez, “Elon Brings One of America’s Most Prominent Nazis Back to Twitter,” Rolling Stone, Dec. 2, 2022, https://www.rollingstone.com/politics/politics-news/elon-musk-twitter-reinstates-neo-nazi-andrew-anglin-account-1234640390/
Daniel Villareal, “Twitter Won't Lift Ban on GOP House Candidate and 'Proud Islamophobe,'” Newsweek, Aug. 19, 2020, https://www.newsweek.com/twitter-wont-lift-ban-gop-house-candidate-proud-islamophobe-1526321
Shanti Das, “Inside the Violent, Misogynistic World of TikTok’s New Star, Andrew Tate,” The Guardian, Aug. 6, 2022, https://www.theguardian.com/technology/2022/aug/06/andrew-tate-violent-misogynistic-world-of-tiktok-new-star
Martin Pengelly, “Far-Right Influencer Known as ‘Baked Alaska’ Sentenced Over Capitol Attack,” The Guardian, Nov. 10, 2021, https://www.theguardian.com/us-news/2023/jan/10/baked-alaska-anthime-gionet-sentenced-capitol-attack
Jordan Valinsky, “Newsmax Reporter Permanently Banned from Twitter for Posting COVID Misinformation,” CNN, Nov. 10, 2021, https://www.cnn.com/2021/11/10/media/newsmax-twitter-emerald-robinson-banned/index.html
Rebecca Klar, “Meta Rejects Recommendation to Suspend Former Cambodian Prime Minister,” The Hill, Aug. 29, 2023, https://thehill.com/policy/technology/4176843-meta-rejects-recommendation-to-suspend-former-cambodian-prime-minister/
In the week after Twitter banned Trump’s account, misinformation about election fraud dropped 73 percent. See Edward Moyer, “After Twitter Banned Trump, Misinformation Plummeted, Report Says,” CNET, Jan. 16, 2021, https://www.cnet.com/news/politics/after-twitter-banned-trump-misinformation-plummeted-says-report/
Oliver Darcy, “Hate Speech Dramatically Surges on Twitter Following Elon Musk Takeover, New Research Shows,” CNN, Dec. 12, 2022, https://www.cnn.com/2022/12/02/tech/twitter-hate-speech/index.html
Taylor Lorenz, “Musk’s Twitter No Longer Bans COVID Misinformation,” The Washington Post, Nov. 29, 2022, https://www.washingtonpost.com/technology/2022/11/29/twitter-covid-misinformation-policy/; Cat Zakrzewski, “Twitter Dissolves Trust and Safety Council,” The Washington Post, Dec. 12, 2022, https://www.washingtonpost.com/technology/2022/12/12/musk-twitter-harass-yoel-roth/; Elon Musk tweet, Twitter, Nov. 24, 2022, https://twitter.com/elonmusk/status/1595869526469533701
Hayden Field and Jonathan Vanian, “Tech Layoffs Ravage the Teams That Fight Online Misinformation and Hate Speech,” CNBC, May 26, 2023, https://www.cnbc.com/2023/05/26/tech-companies-are-laying-off-their-ethics-and-safety-teams-.html#:~:text=Twitter%20effectively%20disbanded%20its%20ethical,safety%20department%2C%20according%20to%20reports
Dan Milmo, “Elon Musk Offers General Amnesty to Suspended Twitter Accounts,” The Guardian, Nov. 24, 2022, https://www.theguardian.com/technology/2022/nov/24/elon-musk-offers-general-amnesty-to-suspended-twitter-accounts
Julia Cohen, “New Twitter, Now With More Hate,” USC Viterbi School of Engineering, April 20, 2023, https://viterbischool.usc.edu/news/2023/04/new-twitter-now-with-more-hate/
TrustLab, “Code of Practice on Disinformation,” Sept. 22, 2023, https://disinfocode.eu/wp-content/uploads/2023/09/code-of-practice-on-disinformation-september-22-2023.pdf
Id.
Ben Goggin, “Verified Accounts Spread Fake News Release About a Biden $8 Billion Aid Package to Israel,” NBC News, Oct. 8, 2023, https://www.nbcnews.com/tech/internet/verified-accounts-spread-fake-news-release-biden-8-billion-aid-package-rcna119372
Vittoria Elliott and David Gilbert, “Elon Musk’s Main Tool for Fighting Disinformation on X Is Making the Problem Worse, Insiders Claim,” WIRED, Oct. 17, 2023, https://www.wired.com/story/x-community-notes-disinformation/
Glenn Chapman, “A Year After Musk's Twitter Takeover, X Remains Mired in Turmoil,” Techxplore, Oct. 26, 2023, https://techxplore.com/news/2023-10-year-musk-twitter-takeover-mired.html
Clare Duffy, “More Than Half of Twitter’s Top 1,000 Advertisers Stopped Spending on Platform, Data Show,” CNN, Feb. 10, 2023, https://www.cnn.com/2023/02/10/tech/twitter-top-advertiser-decline/index.html
Meta, “Introducing Threads: A New Way to Share With Text,” Meta Newsroom, July 5, 2023, https://about.fb.com/news/2023/07/introducing-threads-new-app-text-sharing; Christina Newberry, “34 Instagram Stats Marketers Need to Know in 2023,” Hootsuite, Jan. 24, 2023, https://blog.hootsuite.com/instagram-statistics (“ … more than 2 billion now use the platform monthly, according to the latest Meta earnings call”).
Jay Peters and Jon Porter, “Instagram’s Threads Surpasses 100 Million Users,” The Verge, July 10, 2023, https://www.theverge.com/2023/7/10/23787453/meta-instagram-threads-100-million-users-milestone
Alex Hern, “Zuckerberg’s Meta to Lay Off Another 10,000 Employees,” The Guardian, March 14, 2023, https://www.theguardian.com/technology/2023/mar/14/mark-zuckerberg-meta-layoffs-hiring-freeze
Katie Paul, “Facebook Owner Meta Slashes Business Teams in Final Round of Layoffs,” Reuters, May 24, 2023, https://www.reuters.com/technology/facebook-owner-meta-starts-final-round-layoffs-2023-05-24
Naomi Nix, “Meta Is Done Moderating. On Threads, Users Decide What They See,” The Washington Post, July 14, 2023, https://www.washingtonpost.com/technology/2023/07/14/threads-algorithm-content-moderation
Christianna Silva, “Threads Is Finally Adding a Search Function,” Mashable, Aug. 31, 2023, https://mashable.com/article/threads-search-keyword; Taylor Lorenz, “Threads Blocks Searches Related to COVID and Vaccines as Cases Rise,” The Washington Post, Sept. 11, 2023, https://www.washingtonpost.com/technology/2023/09/11/threads-covid-coronavirus-searches-blocked/
Free Press, “How Tech Companies & Policymakers Can Fight AI Harms & Safeguard Digital Civil Rights,” Sept. 12, 2023, https://www.freepress.net/policy-library/how-tech-companies-policymakers-can-fight-ai-harms-safeguard-digital-civil-rights
Free Press, “Musk Meets with Free Press and Civil-Rights Groups to Discuss Twitter Community Standards, Election Integrity and Content Moderation,” Nov. 2, 2022, https://www.freepress.net/news/press-releases/musk-meets-free-press-and-civil-rights-groups-discuss-twitter-community-standards
Hayden Field and Jonathan Vanian, “Tech Layoffs Ravage the Teams that Fight Online Misinformation and Hate Speech,” CNBC, May 26, 2023, https://www.cnbc.com/2023/05/26/tech-companies-are-laying-off-their-ethics-and-safety-teams-.html
Sheera Frankel, Adam Satariano and Ryan Mac, “Meta Lays Off More Than 11,000 Employees,” The New York Times, Nov. 9, 2022, https://www.nytimes.com/2022/11/09/technology/meta-layoffs-facebook.html
Max Zahn, “How Twitter's $8 Verification Plan Works and Why It's Facing Criticism,” ABC News, Nov. 9, 2022, https://abcnews.go.com/Business/twitters-verification-plan-works-facing-criticism/story?id=92794983
Elon Musk tweet, Twitter, Nov. 19, 2022, https://twitter.com/elonmusk/status/1594131768298315777
Elon Musk tweet, Twitter, Nov. 24, 2022, https://twitter.com/elonmusk/status/1595869526469533701
Taylor Lorenz, “Twitter Ends Its Ban on COVID Misinformation,” The Washington Post, Nov. 29, 2022, https://www.washingtonpost.com/technology/2022/11/29/twitter-covid-misinformation-policy/
Cat Zakrzewski, “Twitter Dissolves Trust and Safety Council,” The Washington Post, Dec. 12, 2022, https://www.washingtonpost.com/technology/2022/12/12/musk-twitter-harass-yoel-roth/
Paul Sawers, “Google Parent Alphabet Cuts 6% of Its Workforce, Impacting 12,000 People,” TechCrunch, Jan. 20, 2023, https://techcrunch.com/2023/01/20/google-parent-alphabet-cuts-6-of-its-workforce-impacting-12000-people/
Nick Clegg, “Ending Suspension of Trump’s Accounts With New Guardrails to Deter Repeat Offenses,” Meta, Jan. 25, 2023, https://about.fb.com/news/2023/01/trump-facebook-instagram-account-suspension/
Thomas Brewster and Richard Nieva, “Google Cuts Company Protecting People from Surveillance to a ‘Skeleton Crew,’ Say Laid Off Workers,” Forbes, Feb. 3, 2023, https://www.forbes.com/sites/thomasbrewster/2023/02/02/jigsaw-google-alphabet-layoffs/?sh=1bcd880c2d71
Kate Conger, Ryan Mac and Mike Isaac, “In Latest Round of Job Cuts, Twitter Is Said to Lay Off at Least 200 Employees,” The New York Times, Feb. 26, 2023, https://www.nytimes.com/2023/02/26/technology/twitter-layoffs.html
Hayden Field and Jonathan Vanian, “Tech Layoffs Ravage the Teams that Fight Online Misinformation and Hate Speech,” CNBC, May 26, 2023, https://www.cnbc.com/2023/05/26/tech-companies-are-laying-off-their-ethics-and-safety-teams-.html
BBC, “YouTube Reinstates Donald Trump's Channel,” March 17, 2023, https://www.bbc.com/news/technology-64993603
Adi Robertson, “Twitter Has Banned Misgendering or ‘Deadnaming’ Transgender People,” The Verge, Nov. 27, 2018, https://www.theverge.com/2018/11/27/18113344/twitter-trans-user-hateful-content-misgendering-deadnaming-ban
Twitter, “Freedom of Speech, Not Reach: An Update on Our Enforcement Philosophy,” April 17, 2023, https://blog.twitter.com/en_us/topics/product/2023/freedom-of-speech-not-reach-an-update-on-our-enforcement-philosophy
Clare Duffy, “Meta’s Latest Round of Layoffs Is Underway,” CNN, April 19, 2023, https://www.cnn.com/2023/04/19/tech/meta-tech-team-layoffs-begin/index.html
Alex Heath, “TikTok’s Head of US Trust and Safety Is Leaving,” The Verge, May 2, 2023, https://www.theverge.com/2023/5/2/23708534/tiktok-head-of-us-trust-and-safety-eric-han-leaves-company
Kali Hays, “Elon Musk Has Chopped Twitter Down to About 1,000 Employees,” Business Insider, May 2, 2023, https://www.businessinsider.com/elon-musk-chops-twitter-down-1000-employees-2023-5
Thomas Maxwell and Hugh Langley, “Google Is Downsizing Its Contract Workforce That Supports YouTube Shortly After One Contractor Team's Union Victory,” Business Insider, May 15, 2023, https://www.businessinsider.com/google-downsizes-contract-workforce-youtube-union-win-2023-5
Amanda Silberling, “Meta Conducts Yet Another Round of Layoffs,” TechCrunch, May 24, 2023, https://techcrunch.com/2023/05/24/meta-conducts-yet-another-round-of-layoffs/
YouTube, “An Update on Our Approach to US Election Misinformation,” June 2, 2023, https://blog.youtube/inside-youtube/us-election-misinformation-update-2023/
Mary Kekatos, “Health Experts Worry as Meta Rolls Back Some COVID Misinformation Policies,” ABC News, June 19, 2023, https://abcnews.go.com/US/health-experts-worry-meta-rolls-back-covid-misinformation/story?id=100192537
Meta, “Introducing Threads: A New Way to Share With Text,” July 5, 2023, https://about.fb.com/news/2023/07/introducing-threads-new-app-text-sharing/
Jordan Valinsky, “Elon Musk Rebrands Twitter as X,” CNN, July 24, 2023, https://www.cnn.com/2023/07/24/tech/twitter-rebrands-x-elon-musk-hnk-intl/index.html
Elon Musk tweet, Twitter, Aug. 18, 2023, https://twitter.com/elonmusk/status/1692558414105186796
Naomi Nix and Sarah Ellison, “Following Elon Musk’s Lead, Big Tech is Surrendering to Disinformation,” The Washington Post, Aug. 25, 2023, https://www.washingtonpost.com/technology/2023/08/25/political-conspiracies-facebook-youtube-elon-musk/
Clare Duffy and Brian Fung, “X Will Allow Political Ads Again and Hire for Safety and Election Teams Ahead of 2024 Elections,” CNN, Aug. 29, 2023, https://www.cnn.com/2023/08/29/tech/x-twitter-2024-election-ads/index.html
Sarah Perez, “YouTube is Giving Creators Violating Policies a Way to Wipe Out Their Warnings,” TechCrunch, Aug. 29, 2023, https://techcrunch.com/2023/08/29/youtube-is-giving-creators-violating-policies-a-way-to-wipe-out-their-warnings/
Tonya Riley, “Twitter, Now X, Will Begin Collecting Users’ Biometric Data,” CyberScoop, Aug. 31, 2023, https://cyberscoop.com/twitter-x-user-biometric-data/
Vittoria Elliott, “Meta Isn’t Enforcing Its Own Political Ads Policy, While the 2024 US Election Looms,” WIRED, Sept. 1, 2023, https://www.wired.com/story/meta-prageru-advertising/
Justin Baragona, “Elon Musk Ponders Twitter Poll on Banning ADL From ‘Free Speech’ Site,” The Daily Beast, Sept. 2, 2023, https://www.thedailybeast.com/elon-musk-ponders-twitter-poll-on-banning-adl-from-free-speech-site-after-bantheadl-trends
Guardian staff and agencies, “Elon Musk’s X Sues California Over New Social Media Transparency Laws,” The Guardian, Sept. 8, 2023, https://www.theguardian.com/us-news/2023/sep/08/elon-musk-twitter-lawsuit-california-free-speech
Taylor Lorenz, “Threads Blocks Searches for ‘Covid’ and ‘Long-Covid,’” The Washington Post, Sept. 11, 2023, https://www.washingtonpost.com/technology/2023/09/11/threads-covid-coronavirus-searches-blocked/
Louise Matsakis, “Google Lays Off Hundreds on Recruiting Team,” Semafor, Sept. 13, 2023 https://www.semafor.com/article/09/13/2023/google-lays-off-hundreds-on-recruiting-team
Sara Fischer, “Musk Says X Will Charge Everyone to Use the Platform,” Axios, Sept. 19, 2023, https://www.axios.com/2023/09/19/musk-x-twitter-charge-all-users-monthly-subscription-fees
Michael Liedtke, “Google Brings Its AI Chatbot Bard into Its Inner Circle, Opening Door to Gmail, Maps, YouTube,” Associated Press, Sept. 19, 2023, https://apnews.com/article/google-artificial-intelligence-bard-gmail-youtube-maps-1229638b82d19afb5226c913821fa1ad
Byron Kaye, “Musk's X Disabled Feature for Reporting Electoral Misinformation — Researcher,” Reuters, Sept. 26, 2023, https://www.reuters.com/technology/musks-x-disabled-feature-reporting-electoral-misinformation-researcher-2023-09-27/
Sarah Perez, “YouTube Relaxes Advertiser-Friendly Guidelines Around Controversial Topics, Like Abortion, Abuse and Eating Disorders,” TechCrunch, Sept. 26, 2023, https://techcrunch.com/2023/09/26/youtube-relaxes-advertiser-friendly-guidelines-around-controversial-topics-like-abortion-abuse-and-eating-disorders/
Erin Woo, “Musk’s X Cuts Half of Election Integrity Team After Promising to Expand It,” The Information, Sept. 27, 2023, https://www.theinformation.com/articles/musks-x-cuts-half-of-election-integrity-team-after-promising-to-expand-it
Mike Clark, “Privacy Matters: Meta’s Generative AI Features,” Meta, Sept. 27, 2023, https://about.fb.com/news/2023/09/privacy-matters-metas-generative-ai-features/
Alex Kirshner, “Twitter Was for News,” Slate, Oct. 5, 2023, https://slate.com/technology/2023/10/elon-musk-x-twitter-news-links-headlines-why.html
Divya Bhati, “Meta to Announce Another Round of Layoffs, Will Fire Employees in Metaverse Silicon Unit,” India Today, Oct. 4, 2023, https://www.indiatoday.in/technology/news/story/meta-to-announce-another-round-of-layoff-will-fire-employees-in-metaverse-silicon-unit-2444214-2023-10-04
Ryan Mac and Cecilia Kang, “Whistle-Blower Says Facebook ‘Chooses Profits Over Safety’,” The New York Times, Oct. 3, 2021, https://www.nytimes.com/2021/10/03/technology/whistle-blower-facebook-frances-haugen.html
Musk filed a third lawsuit targeting another research nonprofit, Media Matters for America, following the research period for this report. He filed this lawsuit on Nov. 20, 2023, claiming the organization manufactured a report showing advertisers’ posts beside white-nationalist and neo-Nazi posts to “drive advertisers from the platform and destroy X Corp.” See Barbara Otutay, “Musk’s X Sues Liberal Advocacy Group Media Matters Over Its Report on Ads Next to Hate Groups’ Posts,” Associated Press, Nov. 21, 2023, https://apnews.com/article/elon-musk-media-matters-lawsuit-advertising-neonazi-1fe499daa600f513af27ffa68d2e8b91
Bryan Pietsch, “Twitter, Now X, Sues Group That Researched Hate Speech on Platform,” The Washington Post, Aug. 1, 2023, https://www.washingtonpost.com/technology/2023/08/01/twitter-lawsuit-center-for-countering-digital-hate/
Guardian staff and agencies, “Elon Musk’s X Sues California Over New Social Media Transparency Laws,” The Guardian, Sept. 8, 2023, https://www.theguardian.com/us-news/2023/sep/08/elon-musk-twitter-lawsuit-california-free-speech
Id.
Derek Thompson, “The Most Expensive Comment in Internet History?” The Atlantic, Feb. 23, 2018, https://www.theatlantic.com/business/archive/2018/02/hogan-thiel-gawker-trial/554132/
James Warren, “When It Comes to Libel, Donald Trump Is a (Courtroom) Loser,” Poynter, Oct. 25, 2016, https://www.poynter.org/newsletters/2016/heres-why-trump-is-a-courtroom-loser-many-times-over/
Naomi Nix and Joseph Denn, “These Academics Studied Falsehoods Spread by Trump. Now the GOP Wants Answers,” The Washington Post, June 6, 2023, https://www.washingtonpost.com/technology/2023/06/06/disinformation-researchers-congress-jim-jordan/
Id.; Cat Zakrzewksi and Cristiano Lima, “GOP Lawmakers Allege Big Tech Conspiracy,” The Washington Post, Feb. 8, 2023, https://www.washingtonpost.com/technology/2023/02/08/house-republicans-twitter-files-collusion/
Id.
Marietje Schaake, “When It Comes to AI and Democracy, We Cannot Be Careful Enough,” Financial Times, Oct. 2, 2023, https://www.ft.com/content/39b89be7-398a-4167-9eeb-58af97b764f3
Global Project Against Hate and Extremism, “Alarming New Data Shows Americans Are Worried About Violent Attacks, Intimidation at Polls,” Aug. 4, 2022, https://globalextremism.org/post/alarming-new-data-shows-americans-are-worried-about-violent-attacks-intimidation-at-polls/
Samuel Woolley and Mark Kumleben, “At The Epicenter: Electoral Propaganda in Targeted Communities of Color,” Protect Democracy, Nov. 2021, https://protectdemocracy.org/project/understanding-disinformation-targeting-communities-of-color/#secti on- (“In Georgia, African Americans and Hispanic Americans were on the receiving end of sophisticated microtargeting efforts erroneously claiming that then-Senate candidate Raphael Warnock ‘celebrated’ Fidel Castro. In Arizona, Hispanic American and Native American communities faced a cascade of untrue digital messaging over Twitter about the voting process. In Wisconsin, multiple communities of color from Madison to Milwaukee were targeted with lies about mail-in ballot fraud and ballot dumping.” (internal citations omitted)
Oversight Board, “Oversight Board Publishes Policy Advisory Opinion on Meta’s Cross-Check Program,” December 2022, https://oversightboard.com/news/501654971916288-oversight-board-publishes-policy-advisory-opinion-on-meta-s-cross-check-program/
Free Press, “Facebook Targeted by Mobile Billboard Circling Capitol Hill Demanding That Company Close the Trump Ad Loophole,” Sept. 30, 3021, https://www.freepress.net/news/press-releases/mobile-billboard-circling-capitol-hill-demands-facebook-close-trump-ad-loophole
Melissa Quinn, “Trump's Social Media Attacks Bring Warnings of Potential Legal Consequences,” CBS News, Aug. 25, 2023, https://www.cbsnews.com/news/trump-bail-conditions-social-media-legal-case-trial/
Meta has an exhaustive Ad Library that it launched in 2018. Experts value the tool since it was the first of its kind. But it’s largely an opaque and clunky database that makes it difficult for users to understand the breadth of trends in political ads or to parse through other metrics to determine why they or others encounter specific content. Rob Leathern, “Expanded Transparency and More Controls for Political Ads,” Meta, Jan. 9, 2020, https://about.fb.com/news/2020/01/political-ads/
Meta’s Oversight Board has also recommended that the company track core metrics and enhance transparency, both of which it has so far failed to do at scale. See Oversight Board, “Oversight Board Publishes Policy Advisory Opinion on Meta’s Cross-Check Program,” December 2022, https://oversightboard.com/news/501654971916288-oversight-board-publishes-policy-advisory-opinion-on-meta-s-cross-check-program/
European Commission, “The Enforcement Framework Under the Digital Services Act,” https://digital-strategy.ec.europa.eu/en/policies/dsa-enforcement
Jennifer Saba, “Elon Musk’s X Is a Black Hole of Value,” Reuters, Oct. 3, 2023, https://www.reuters.com/breakingviews/elon-musks-x-is-black-hole-value-2023-10-03/

How Social-Media Rollbacks Endanger Democracy Ahead of the 2024 Elections

Big Tech Backslide

TOPLINE RECOMMENDATIONS

Litigation to Chill Tech Accountability

Mass Layoffs Make Content Moderation Harder

Policy Rollbacks Deprioritize User Safety & Platform Integrity

Executive Summary

Big tech's empty promises

Understanding the Nature of Platform Rollbacks

Big Tech backslide

Reinstated Trump

Laid Off Content Moderators And/Or Trust & Safety Teams

Weakened Privacy & AI Policies

Stopped Moderating Covid Lies

Weakened Political-Ad Policies

Meta Policy Rollbacks

Imposed User Fact-Checking Limits

Reinstated Trump

Laid Off Content Moderators And/Or Trust & Safety Teams

Weakened Privacy & AI Policies

Stopped Moderating Covid Lies

Weakened Political-Ad Policies

Twitter Policy Rollbacks

Imposed User Fact-Checking Limits

Policy Rollbacks Jeopardize Platform Integrity

Mass Layoffs Erode Platform Functionality & Moderation

Is TikTok checking all the boxes?

Reinstatement & Monetization of Dangerous and Extremist Accounts

A League of Its Own: Twitter’s Total Failure

Threads Launch Spreads Meta Too Thin

AI to the Rescue?

A Year of Backsliding: Timeline

Cross-Sector Attacks on Tech Accountability

Strategic Lawsuits Against Platform Transparency

The forecast ahead

What Governments MUST Do

What Platforms Must Do

Recommendations for 2024

Vacation Days

Pet Lovers

Methodology & Acknowledgements

Endnotes:

How Social-Media Rollbacks Endanger Democracy Ahead of the 2024 Elections

Big Tech Backslide

TOPLINE RECOMMENDATIONS

Litigation to Chill Tech Accountability

Mass Layoffs Make Content Moderation Harder

Policy Rollbacks Deprioritize User Safety & Platform Integrity

Executive Summary

Big tech's empty promises

Understanding the Nature of Platform Rollbacks

Big Tech backslide

Reinstated Trump

Laid Off Content Moderators And/Or Trust & Safety Teams

Weakened Privacy & AI Policies

Stopped Moderating Covid Lies

Weakened Political-Ad Policies

Meta Policy Rollbacks

Imposed User Fact-Checking Limits

Reinstated Trump

Laid Off Content Moderators And/Or Trust & Safety Teams

Weakened Privacy & AI Policies

Stopped Moderating Covid Lies

Weakened Political-Ad Policies

Twitter Policy Rollbacks

Imposed User Fact-Checking Limits

Policy Rollbacks Jeopardize Platform Integrity

Mass Layoffs Erode Platform Functionality & Moderation

Is TikTok checking all the boxes?

Reinstatement & Monetization of Dangerous and Extremist Accounts

A League of Its Own: Twitter’s Total Failure

Threads Launch Spreads Meta Too Thin

AI to the Rescue?

A Year of Backsliding: Timeline

Cross-Sector Attacks on Tech Accountability

Strategic Lawsuits Against Platform Transparency

The forecast ahead

What Governments MUST Do

What Platforms Must Do

Recommendations for 2024

How Social-Media Rollbacks Endanger Democracy
Ahead of the 2024 Elections

How Social-Media Rollbacks Endanger Democracy
Ahead of the 2024 Elections