The Cost of Weaker Moderation
CCDH analysed 16,164 Facebook and Instagram comments about the EU Return Regulation vote and classified 43% as hateful or anti-migrant; 1,532 comments contained hate and 251 of those contained gendered abuse.
Executive summary
AI-generatedThis summary was generated by AI from the original report to make it easier to scan and cite. It is not a substitute for the source. Read the original above.
CCDH examined 16,164 comments beneath four Facebook and Instagram posts by MEPs Abir Al-Sahlani and Sebastian Tynkkynen following the European Parliament vote on the EU Return Regulation. Researchers classified 6,986 comments, or 43%, as hateful or anti-migrant.
The study found 1,532 hate comments, including threats, dehumanising language and attacks based on protected characteristics. Of these, 251 contained gendered abuse. A further 5,454 comments were anti-migrant, and 2,673 used a variation of the phrase “send them back.”
CCDH downloaded the comments with a third-party tool and used GPT-4.1 for classification, validating the model against 500 human-labelled comments and reporting an F1 score of 0.85. The report argues that Meta’s weakened moderation policies are allowing violent abuse to flourish and calls on the European Commission to examine the changes under the Digital Services Act.
Think this summary is wrong? Contact us.