Meta Oversight Board finds account bans lack due process, calls for clearer rules

Meta's Oversight Board says account bans lack due process, transparency

Meta's Oversight Board, the independent governing body that makes policy recommendations to the tech company, said Thursday that Meta's account deactivations lack due process, violations are doled out without clarity, and there's little customer support for appeals. The board, which recently received increased funding to continue its work through 2028, launched the investigation into Meta's account violations policy earlier this year to review a case involving threats of violence against a journalist. The board agreed that Meta was right to permanently disable the account due to the severity of the threats. However, in looking into the matter, the board found what it described as "systemic human rights concerns," and a "lack of transparency and consistency" when it came to Meta's two-system approach to disabling accounts. That's a reference to the two different ways accounts can be actioned against: one that involves strikes, some that can be severe, and another for "egregious" violations that merit permanently disabling an account. The board said the difference between what leads to one type of violation or the other isn't clear or well-documented. The board also called out the social media giant for charging users for Meta Verified access, which is supposed to include "24/7 access to email or chat agent support," but failing to provide users with disabled accounts any sort of "meaningful assistance." Meta users across Facebook, Instagram, and other Meta apps have dealt with this problem for years. As Meta's moderation systems became more automated over the years, there has been little recourse when they make mistakes to get any sort of help from the company. This has been devastating to users who have lost their personal accounts, or even those belonging to their business, over false accusations. Some have even filed lawsuits over the matter or are in the process of doing so. The board took up the matter of permanent bans in a landmark case after a wave of recent bans that affected Facebook and Instagram No support, no appeal process After covering the bans, TechCrunch received a constant stream of reports from impacted users, begging us to bring their cases to Meta's attention For instance, retired L.A. County firefighter and paramedic, Richard Pauwels, who was building a wellness brand on Meta's platforms, claimed his personal account was banned without any specific post being identified and without human review. Automated alleged child sexual exploitation (CSE) violations are another common occurrence leading to bans, which are devastating to innocent parties. For instance, a PR professional reached out to us, asking to remain anonymous because their account was banned for false accusations of CSE, despite no content being cited in their ban and no new posts from them in weeks. Meta's allegation was "evil and vile," they told us, and filed a case with the Oversight Board. Another person, Manomi Jayakody, said their account was also banned for CSE, again without any specific content, action, or violation being flagged. "I fully understand and support the critical importance of CSE enforcement and online safety. However, when accounts are flagged under such serious categories without due process, transparency, or consistent human oversight ... the consequences for innocent users are severe," they wrote to TechCrunch and other reporters in an email. "In my case, I received no explanation, no evidence, and no opportunity to clarify or correct any alleged issue." In another example, a bird rescue operation, of all things, was banned for CSE material. "It's an account about our rescue pigeons. It has over 60K followers, and we regularly use that account to contact other volunteers who rescue birds and find adopters or medical assistance for all the birds that we help," they pleaded in an email to TechCrunch. Another content creator, Albert Olgaard, with 325,000 Instagram followers, said his Meta accounts for his business were shut down overnight for allegations of "fraud," with no citations of what he did wrong. When he tried to appeal, a message on the screen stated, "You cannot request another review of this decision." The move financially impacted him, he said. Designer Alex Smola also woke up one day to find his account permanently disabled. "The account was tied to legitimate business activity, including page access, client communication, marketing, advertising, and lead generation," Smola wrote in an email to TechCrunch. "Since the deletion, I have suffered ongoing business disruption, reputational harm, and significant stress, while being unable to obtain any meaningful human review through Meta's support channels." That mirrors many other complaints TechCrunch fielded from Meta Verified subscribers who said they didn't get the support they paid for through the subscription. One person even said the Meta agent ended their chat when they referenced the EU regulation, GDPR. These are just a smattering of the reports we've received since covering the topic, and the inbound emails continue from people desperate for help. Meta's Board highlights lack of due process As exemplified by these examples and others, there are due process concerns around account bans. As a result, the board is recommending that Meta offer users a dashboard where they can easily review their account stats, past violations, and information on appeal options, along with clear notifications about the violations at the time they are imposed. These warnings or violation reports should include the time they were imposed, the specific rule that was violated, the sanction imposed, and appeal options. Additionally, the board says Meta should provide users with information on the role AI plays in content review and the imposition of warnings or penalties, among other things. In response to the board's report, Meta shared the following statement: "We welcome the Oversight Board's decision on this case. The Board upheld Meta's decision to permanently disable an account. After conducting a review of the recommendations provided by the Board, we will update this post with initial responses to those recommendations."

Meta Oversight Board Calls for Clearer Account Ban Rules

The recommendation came in a case involving Meta's permanent disabling of an Instagram account with more than 70,000 followers after it posted threats against a female journalist and repeatedly violated platform rules. While the Board upheld the account's removal, it used the decision to scrutinise Meta's broader approach to account governance, transparency and appeals. The Board described the matter as its "first case assessing the permanent disabling of a user's account". The decision also highlighted widespread user frustration with Meta's enforcement systems. The Board received more than 750 public comments as submitted in the case, while noting it had received "innumerable complaints" about disabled accounts since it began taking cases in 2020. Users said they often did not understand why their accounts had been disabled, could not access effective appeals, could not download their content, or believed that decisions were made automatically without meaningful human review. Issues flagged by the Oversight Board: * Delayed action on violent threats: The Board said it was "seriously concerned that Meta did not review either of these clear and credible threats swiftly when they were posted, delaying removal and exposing the targeted journalist to intolerable risk for a prolonged period." * Inadequate recourse for targets: According to the decision, the journalist had to rely on personal contacts within Meta to trigger action. The Board noted that "any user in such a situation without similar connections would be faced with no recourse." * Due process concerns for banned users: The Board said delayed enforcement also harmed due process for the account holder because earlier intervention could have provided "more opportunities to understand the nature of his wrongdoing and correct his behavior." * Confusing and contradictory rules: The Board found Meta's account-governance framework difficult to follow across multiple policy pages and noted that "some of the information in these sources is contradictory." It cited one example in which the Disabling Accounts page said users could receive a 30-day content-creation restriction after five strikes, while the Restricting Accounts page stated that the same penalty would apply after 10 strikes. * Lack of transparency around Instagram penalties: The Board said Meta "does not publicly set out all the restrictions it can apply to Instagram accounts for violations before and up to disabling them." * No temporary suspensions on Instagram: Meta told the Board that it "does not temporarily suspend Instagram accounts in response to policy violations." Instead, it mainly restricts users from going live before eventual disablement. * No clear framework for permanent bans: The Board warned that Meta lacks "a clear framework guiding decisions to permanently disable an account for 'egregious' safety concerns." Clear rules and account-status information: A "clear and comprehensive guidance" on account-disablement rules was recommended. Specifically, platforms should: * Explain which violations trigger graduated enforcement and which can result in immediate disablement. * Publicise thresholds for escalating penalties, including permanent disablement. * Clearly distinguish between different categories of enforcement, such as regular and severe strikes. * Explain "policy frameworks and factors internal enforcement teams consider when permanently disabling accounts." The Board also said users should have access to dashboards showing current account status, past violations, appeal options and pending appeals. This information should be downloadable and remain accessible even after permanent disablement. The decision additionally noted that "Meta's expansion of the role of AI assistants on its platforms may help better explain rules to users at key moments," while stressing that policy documents and AI-generated explanations should remain consistent. Government requests, automation and notifications: Users should receive "clear, prominent and timely notifications" explaining the rule violated, penalties imposed and appeal options. It further recommended that platforms disclose: * "Information on the role of any government request for either the review or disabling of an account." * "Information on the role of automation in the review of content or behavior and the imposition of warnings or penalties." Cross-platform coordination on violent threats: The Board recommended that social media companies create a programme to share information about accounts that "credibly threaten serious violence." Pointing to existing industry initiatives such as the Global Internet Forum to Counter Terrorism and the Lantern programme on child safety, it said the new mechanism should: * Share threat information across platforms, including details on targets and threat sources. * Establish dedicated channels for "journalists and human rights defenders." * Share best practices for assessing threats, preserving evidence and making referrals to law enforcement where appropriate. Appeals and human review: Appeals systems should allow users to submit written explanations and continue to access the policy justification for enforcement decisions. It also warned against appeal processes that merely rerun automated decisions. "Appeals of account disablement are not effective if they run the same decision through the same automated process, expecting different results with the same inputs," the decision said. While supporting greater use of AI, the Board said "prioritizing of human review in edge cases will be important," particularly for permanent account-disablement decisions. Transparency reporting: Finally, the Board called for detailed transparency reports containing: * The total number of accounts disabled for policy violations. * A breakdown between accounts disabled under "'one and done' rules" and those removed through graduated enforcement systems such as strikes. * The specific rules that led to disablement. * The number of accounts disabled following government or law-enforcement review requests. * The number of disabled accounts referred to law enforcement. * Regional and language-wise breakdowns of enforcement data. Why this matters for India: The Oversight Board's recommendations arrive as India increases both the pace and scope of online content regulation. Earlier this year, the government shortened compliance timelines for intermediaries responding to takedown orders to three hours, while proposed amendments to the IT Rules would increase oversight of user-generated "news and current affairs" content and more closely link compliance with official directions to safe-harbour protections. At the same time, Meta has reportedly included India among a small group of countries in which flagged content can be automatically restricted upon government requests. Recent enforcement has affected journalists, digital news outlets, satirists and political actors. MediaNama's March 2026 censorship tracker documented more than 40 instances of content blocking, account restrictions and takedowns, including actions against Molitics, National Dastak, Rajeev Nigam and news publications in Kashmir.