European Commission adopts final AI labeling Code of Practice

On 10 June the European Commission published the final Code of Practice that translates Article 50 of the EU AI Act into concrete steps for labeling AI‑generated or manipulated content.

The voluntary playbook, drafted by six independent experts with input from more than 187 participants, becomes operative about seven weeks before the mandatory transparency obligations start on 2 August 2026. From that date deepfakes and AI‑generated text on matters of public interest must carry machine‑readable metadata or imperceptible watermarks, and users must be told when they are interacting with an AI system such as a chatbot.

Non‑compliance can attract fines of up to €15 million or 3 % of worldwide turnover. The guidelines require signed, tamper‑evident metadata, watermarking, and provide free detection tools, with a transitional period for systems already on the market. “Europeans have a right to know whether what they see, hear or read has been made or altered by AI… Transparency is how we protect trust,” said Henna Virkkunen, the Commission’s executive vice‑president for tech sovereignty, security and democracy.