The European Commission has published a voluntary Code of Practice setting out how providers and users of generative artificial intelligence should mark and label AI-generated content ahead of new EU transparency rules due to take effect on 2 August 2026.
The Code is designed to help “providers” (companies that develop generative AI systems) and “deployers” (those that use or integrate them) meet transparency obligations under the EU’s AI Act from that date, the Commission declared in a statement on Wednesday.
Under the AI Act rules cited in the Commission’s statement, deepfakes and AI-generated or AI-manipulated text published on matters of public interest will have to be clearly labelled.
People will also have to be told when they are interacting with an interactive AI system such as a chatbot.
The Commission said the Code was drafted by six independent experts with input from more than 180 stakeholders, including AI providers, deployers, small and medium-sized enterprises, academics, public sector bodies and civil society organisations.
What the Code covers
The Code has two sections, one aimed at providers and one at deployers, the Commission said.
For providers, it sets out how AI-generated or AI-manipulated audio, images, video or text should be marked in a machine-readable way — meaning computers can detect that content has been generated or altered by AI.
For deployers, it explains that deepfakes and AI-generated or AI-manipulated text intended to inform the public on matters of public interest must be clearly labelled where there has been no human review or editorial control.
The Code is open for signatures, and once it is approved as adequate by the Commission and the AI Board, signatories will be able to use it to demonstrate compliance with the relevant AI Act obligations from 2 August 2026.
Henna Virkkunen, the Commission’s Executive Vice-President for Tech Sovereignty, Security and Democracy, said: “Europeans have a right to know whether what they see, hear or read has been made or altered by AI.”
