EU consults on DSA election guidelines that would have platforms label deepfakes and test recommender systems before votes

The European Commission has opened a targeted consultation on revised election guidelines under the Digital Services Act, covering how very large online platforms and search engines should handle generative AI and information manipulation risks. Stakeholders can respond until 23 November 2026.

Under the draft, AI-generated deepfakes should be clearly labelled and keep their label when shared, and illegal deepfakes of candidates should come down fast. Platforms should detect coordinated inauthentic behaviour early, spot suspicious trends before they go viral, test their recommender systems for manipulation before every national election and keep a record of changes made in the six months before a vote.

The draft also asks for dedicated channels for candidates facing threats or impersonation, and for platforms to report to the Commission and national authorities within 30 days of an election. The Commission first adopted the guidelines in April 2024; it says they are content-neutral and do not ask platforms to delete lawful content.