EU AI transparency code timeline: first draft to final, June 2026

Yes, it is final. The Commission's Code of Practice on Transparency of AI-generated Content page lists the final code as published at the closing plenary on 10 June 2026, after drafts on 17 December 2025 and 3 March 2026. The page says the transparency obligations apply from 2 August 2026 and that about 190 companies and organisations had signed by the end of July 2026. This reports the page and is not legal advice.

What were the steps from draft to final?

The timeline on the page runs: September 2025 consultation and call for expression of interest; 5 November 2025 kick-off plenary; 17 December 2025 first draft; 3 March 2026 second draft; 8 May 2026 draft guidelines published for consultation; and 10 June 2026 closing plenary with the final code. The code was drafted by independent experts in a multi-stakeholder process the AI Office facilitated, in two working groups, providers and deployers.

Has the Commission confirmed the code is adequate?

Yes, per the page: the Commission and the AI Board have confirmed it is an adequate voluntary tool to demonstrate compliance with Article 50(2), (4) and (5). Signatories can rely on it; others show their own measures are adequate.

Does the grace period change anything for video made now?

The FAQ says Article 50 applies from 2 August 2026 and that a limited grace period covers only AI systems placed on the market before that date, and only the Article 50(2) marking and detection duty, which then applies from 2 December 2026. Content generated before 2 August 2026 does not need to be labelled retroactively, though the Commission encourages it where possible.

For files you generate now, keep a record per output. Sume's docs on Jobs and results say to store the job id from the submit response, which gives you one stable key per generated file.