Major social media platforms are facing significant challenges with their content moderation practices in Europe, according to new data from an independent appeals body established under European Union regulations. The Appeals Centre Europe, based in Dublin, has documented thousands of moderation errors across platforms including TikTok, YouTube, Facebook, and Instagram during its first year of operation.
The center processed more than 24,000 complaints between April 2025 and March 2026, with approximately 12,000 cases qualifying for formal review. The volume of eligible disputes increased dramatically throughout the year, reaching nine times the initial monthly levels by March 2026.
Hate speech moderation emerged as a particularly problematic area. When examining over 1,400 cases where platforms had declined to remove reported content, independent reviewers reversed those decisions in 70% of instances. TikTok showed the highest reversal rate at 83%, followed by Instagram at 74%, Facebook at 61%, and YouTube at 58%.
The reviewed content included antisemitic posts, anti-migrant rhetoric, racist abuse, anti-Roma speech, and attacks targeting LGBTQI+ communities. Thomas Hughes, chief executive of the appeals center, emphasized that online hate and harassment have tangible real-world impacts on affected individuals and communities.
Beyond hate speech, platforms struggled with identifying scams, manipulated media, and misinformation. One case involved an AI-generated TikTok video about the Russia-Ukraine conflict that featured a fabricated voiceover overlaid on authentic news footage. Despite violating platform policies, the content remained accessible weeks after review.
Another incident involved racist Instagram comments posted following a Liverpool-Galatasaray Champions League match, where Black players were compared to monkeys. While all 20 reviewed comments were found to violate Meta’s hateful conduct policies, several remained visible on the platform.
The appeals process also revealed significant overmoderation issues. Reviewers overturned platform decisions in 52% of cases involving removed content. One example involved a Czech photographer whose artistic nude photograph was removed from Facebook despite having blurred nipples to comply with Meta’s policies. The image was later reinstated with age restrictions following the center’s recommendation.
Account suspensions and restrictions generated the largest category of complaints, with more than 14,000 users seeking reviews. However, platforms frequently failed to provide necessary evidence for independent assessment. By March’s end, sufficient documentation existed to fully evaluate fewer than 150 account-suspension cases, resulting in over 7,000 default decisions favoring users due to platforms’ failure to cooperate.
Geographic participation varied significantly across European Union member states. France generated the highest absolute number of eligible disputes, followed by Belgium, Italy, Spain, and Germany. On a per capita basis, Belgium led, followed by Lithuania, Cyprus, Malta, and Slovakia.
Civil society organizations played a substantial role in filing complaints, particularly regarding hate speech. Groups including Slovakia’s digiQ, Poland’s NEVER AGAIN Association, and Belgium’s Media Diversity Institute Global submitted extensive documentation of racist, antisemitic, and xenophobic content that platforms had failed to remove.
While the center’s decisions lack binding legal force, the Digital Services Act requires platforms to engage with certified dispute-settlement bodies in good faith. Repeated failures to provide evidence or participate meaningfully could attract increased regulatory scrutiny from national authorities and the European Commission.
Hughes noted that many moderation failures appear linked to inadequate contextual review, insufficient language expertise, and users deliberately circumventing automated detection systems through tactics such as replacing slurs with symbols or mixing alphabets.
The establishment of this external appeals mechanism represents a significant shift in how content moderation disputes are handled, providing European users with recourse beyond platforms’ internal systems for challenging decisions about their posts and accounts.
Leave a Reply