Code of Practice on Transparency of AI-generated Content

Summary

A voluntary code that helps providers and deployers of generative AI systems comply with Article 50(2), (4) and (5) of the AI Act. Section 1 sets rules for providers on machine-readable marking of AI-generated or manipulated content, with digitally signed metadata and imperceptible watermarking, on detection mechanisms, and on testing and compliance processes. Section 2 sets rules for deployers on labelling deepfakes and AI-generated or manipulated text published on matters of public interest, with EU icons for labelling. Signatories can rely on the code to demonstrate compliance.

Safety topics

Status notes

Drawn up by independent experts in a multi-stakeholder process facilitated by the AI Office, in two working groups (providers and deployers) led by independent chairs and vice-chairs. Drafting began with a kick-off plenary on 5 November 2025; the first draft was published on 17 December 2025 and the second on 3 March 2026, and the final code was published at the closing plenary on 10 June 2026. On 8 July 2026 the Commission concluded that the code adequately covers the obligations in Article 50(2), (4) and (5) of the AI Act, and the following day the AI Board adopted its adequacy assessment; the Commission's opinion was published on 9 July 2026. Adherence to the code does not constitute conclusive evidence of compliance, and providers and deployers that comply by other means must show that their measures are adequate. By the end of July 2026 about 190 organisations had signed the code. Signatories work together in two signatory task forces, on marking and detection and on labelling. The AI Office will consider facilitating formal updates of the code at least every two years. The code is complemented by the Commission's guidelines on Article 50 (EU-19).

Sources

Open the official source https://digital-strategy.ec.europa.eu/en/policies/code-practice-ai-generated-content

Secondary source: https://digital-strategy.ec.europa.eu/en/library/commission-opinion-assessment-code-practice-transparency-ai-generated-content

Local copy

Verification

History

1. 2026-10-17notes: Drafting began with a kick-off plenary on 5 Nov 2025. First draft 17 Dec 2025. S → Drawn up by independent experts in a multi-stakeholder process facilitated by th — the AI Board's adequacy assessment is dated to the day after the Commission's 8 July 2026 conclusion, as the opinion page says (version 1 put it after the 9 July publication); working groups, the duty of non-signatories, the signatory task forces and the update cycle added from the stored pages2026-10-17-recheck-europe-1 · Claude Code (claude-opus-5-5)
2. 2026-10-17summary: A voluntary code that helps providers and deployers of generative AI meet AI Act → A voluntary code that helps providers and deployers of generative AI systems com — the marking techniques (signed metadata, watermarking), the detection and compliance commitments and the deployers' rules now come from the code's own text; reliance on the code to demonstrate compliance added from the policy page2026-10-17-recheck-europe-1 · Claude Code (claude-opus-5-5)
3. 2026-10-17date_last_updated: 2026-07-08 → – — 8 July 2026 is the Commission's adequacy conclusion, not a new version of the code; no stored source shows a revised code, so the field is cleared and the assessment stays in the notes2026-10-17-recheck-europe-1 · Claude Code (claude-opus-5-5)
4. 2026-10-17issuing_body: European Commission (AI Office), drafted by independent chairs; assessed as adeq → European Commission (AI Office) — the field names the institution that issued the code, the Commission's AI Office, which facilitated it; who drafted it and the adequacy assessment are facts for the notes2026-10-17-recheck-europe-1 · Claude Code (claude-opus-5-5)

Related records

EU AI Act European Union · Partially in force

Article 50 transparency guidelines European Union · Published (non-binding, current)

Citation

European Commission (AI Office). Code of Practice on Transparency of AI-generated Content. 2026-06-10. https://digital-strategy.ec.europa.eu/en/policies/code-practice-ai-generated-content. Status: Published (non-binding, current); verified 2026-10-17 (level A). Local copy archive/EU-16/2026-10-17/official.html retrieved 2026-10-17 16:47:14 UTC, SHA-256 2e9d3af6eebbe314184572cbd04644c9ce149e9ee1b95465a288e0eeb3ccebab. Every AI Policy Ever, version 1.18 (snapshot 2026-10-18), record EU-16.

Improve this record

Know of an amendment, a status change or an error? Report news or an error on GitHub; the form opens with this record filled in.