Meta Funded the EU’s Social Media Appeals Body — Then Stonewalled Its Rulings

When the European Union’s Digital Services Act (DSA) carved out a space for independent, out-of-court dispute resolution, the idea was simple: give users a faster, fairer way to fight wrongful bans than navigating slow-moving courts. The Appeals Centre Europe (ACE) was created precisely for this purpose. The irony is that its launch was bankrolled by the very company whose moderation mistakes it now routinely exposes.

Meta — through the Meta Oversight Board Trust — provided a one-off €15 million grant to establish the Dublin-based ACE . The body was certified by Ireland’s media regulator, Coimisiún na Meán, in September 2024 to handle disputes against Facebook, Instagram, TikTok, and YouTube . Since opening its doors in November 2024, it has been flooded with complaints from EU users who feel their content was unfairly removed or their accounts wrongly suspended.

But the relationship between the Appeals Centre and Meta has soured. As the ACE itself puts it, Meta’s platforms “often dragged their feet, refusing to engage with the body sufficiently” . This combination of Meta’s seed funding and subsequent resistance has turned the ACE into a real-world stress test for the DSA’s enforcement capabilities, revealing a glaring gap between the law’s ambition and its actual power.

A 75% Overturn Rate Points to Systemic Failures at Meta

The most damning statistic from the ACE's work is its decision record. The body’s first transparency report, covering cases from November 2024 through August 2025, showed that more than three-quarters of platforms’ original moderation decisions were overturned . This staggering figure includes cases where the ACE disagreed with the platform after reviewing the content, and instances where the platform reversed its own decision once an independent body got involved.

For Meta specifically, an early batch of rulings saw the Appeals Centre side with the user and overturn Meta’s original decision in 77 cases relating to Facebook and Instagram . To be clear: these were not borderline judgment calls. These were decisions where an impartial body of experts concluded that Meta’s automated or human moderation teams had made the wrong call, and that a user’s account or post should never have been penalized in the first place.

This error rate suggests that Meta’s internal moderation apparatus — a system that affects hundreds of millions of European users — is making incorrect decisions on an industrial scale. The appeals body’s most recent annual report noted 24,000 disputes in the period from April 2025 to March 2026 alone, covering issues such as hate speech, account suspensions, adult nudity, misinformation, and fraud .

A Contradictory Posture: Funding, then Foot-Dragging

The most awkward tension in the whole process isn’t the overturned decisions, but Meta’s own posture towards the body. The company’s Oversight Board Trust celebrated the ACE’s launch as a key part of the EU’s new regulatory landscape . But behind the scenes, the cooperation the DSA mandates has been conspicuously absent.

The ACE noted that when Meta did respond to case inquiries, its replies were often too slow or lacked the necessary information to properly adjudicate the dispute . By the end of August 2025, Meta had implemented only about half of the centre’s overturn decisions for Facebook and Instagram — representing roughly 100 cases where the company simply disregarded the body’s conclusion .

YouTube was cited as an even worse offender, often providing no content at all, which made it impossible for the ACE to decide the merits of a case . But Meta’s pattern of partial cooperation and sporadic stonewalling underscores the fundamental vulnerability of the system: the platform can, for all practical purposes, opt out when it’s inconvenient.

The DSA’s Structural Weakness: Non-Binding Rulings

The root of the problem lies in the legal design. Under Article 21 of the DSA, out-of-court dispute settlement bodies like the ACE are empowered to review platform decisions, but their rulings carry no legal force . Platforms are required to “engage in good faith,” a phrase that has proven almost meaningless in practice without a clear enforcement mechanism backing it up.

The European Commission has opened multiple DSA investigations into Meta over issues such as political content demotion, inadequate researcher data access, and failure to protect minors . In April 2026, the Commission issued a preliminary finding that Meta was breaching the DSA by not keeping children under 13 off Instagram and Facebook — an infraction that could eventually lead to a fine of up to 6% of Meta’s global annual turnover .

Across Big Tech, Brussels imposed at least €3.77 billion in fines in 2025 alone, signaling an aggressive new phase of enforcement . But these headline penalties target systemic compliance failures, not individual decisions to ignore a non-binding out-of-court ruling. For a user in Berlin or Barcelona who simply wants their Facebook account reinstated, the lack of bite in the ACE process remains a frustrating dead end.

What This Means for EU Users

The demand for independent review is real. More than 30,000 complaints have been lodged with the Appeals Centre Europe since it began operating in November 2024, proving that people across the EU actively seek alternatives to opaque, automated platform justice . The process is free for users, and it operates with a speed that traditional legal mechanisms cannot match .

But the ACE’s experience also clarifies the limits of the current DSA framework. The body has demonstrated that it can surface systematic moderation errors and create a public record of platform non-compliance. It cannot, however, force Meta — or any other platform — to change its behavior.

Until EU lawmakers grant ODS bodies the authority to issue binding decisions, or until the Commission begins penalizing platforms for bad-faith engagement in these specific proceedings, the system will continue to provide a diagnosis without a cure. It tells users precisely when Meta has wronged them, but leaves the company free to decide whether to make it right.