Code of practice for marking AI-generated content
On 10 June 2026 the European Commission published the final Code of Practice on marking and labelling AI-generated content. The code is voluntary and explains how providers should apply a machine-readable mark to generated audio, images, video and text, and how deployers should label deepfakes and text on matters of public interest.
Why it matters
Until 2 August 2026 the duty to mark existed only in the text of the law; the code spelled it out as concrete steps by which signatories will be able to show compliance once it is approved. An editorial assessment.
Press release IP/26/1328 says the code was written by six independent experts with input from over 180 stakeholders (the Commission's figures). It has two sections. The providers' section concerns machine-readable marking of audio, images, video and text so that they can be detected as artificial. The deployers' section concerns labelling deepfakes and text published to inform the public on matters of public interest where there was no human review or editorial control. The code is 'open for signatures'; signatories will be able to demonstrate compliance once the Commission and the AI Board approve it as adequate. The Commission promises separate guidelines that will clarify the scope of the obligations. What the record does not claim. The code is not mandatory: the duty comes from the AI Act. We did not read the text of the code, the icons or the Q&A pages the press release links to. The identifier 'Regulation 2024/1689, Article 50' that the report gave is the law, not this document.