Final Code of Practice on marking AI-generated content published
Adopted 2026-06-13 · ≈ 2 min read · 
edited by Dirk Baaijen
On 10 June 2026 the European Commission published the final Code of Practice on marking and labelling AI-generated content — the practical instrument under the Article 50 transparency duty, which formally applies from 2 August 2026.
On 10 June 2026 the European Commission published the final Code of Practice on marking and labelling AI-generated content. It is not a new law but the practical instrument for an existing duty: Article 50 of the AI Act, which aims to make synthetic content recognisable. The code arrives deliberately ahead of the application date, so that providers and deployers can prepare.
What the code covers
The code has two parts, tracking the two sides of Article 50:
- Providers (Article 50(2)): anyone placing a generative AI system on
the market marks its output — audio, image, video, text — in a machine-readable format that is detectable, interoperable, robust and reliable. The code gives this concrete shape rather than leaving each provider to invent its own approach.
- Deployers (Article 50(4)): anyone using image, audio or video
material that constitutes a deepfake labels it as artificially generated or manipulated. For this the code provides optional EU icons for a recognisable, uniform label.
Voluntary, but the duty beneath it is not
The code itself is voluntary: it is a streamlined way to demonstrate compliance, not a separate obligation. The underlying Article 50 transparency requirements are not — they are legally binding. Following the code gives a recognisable route to compliance; not following it means demonstrating compliance with Article 50 some other way. In practice such a code often becomes the benchmark against which supervisors and customers measure. It is designed to keep pace with technology: it is reviewed at least every two years.
Timeline
Article 50's transparency obligations formally apply from 2 August 2026, with a transitional period until 2 December 2026 for systems already on the market before that date. The Digital Omnibus agreement of 7 May 2026 shifts several AI Act deadlines; until publication in the Official Journal, 2 August 2026 remains the formal date for Article 50. The code is available now — two months ahead of that date — so the market can get ready.
Sources
- https://digital-strategy.ec.europa.eu/en/policies/code-practice-ai-generated-content
Commission page with the final Code of Practice (10 June 2026): section 1 for providers, section 2 for deployers. - https://digital-strategy.ec.europa.eu/en/faqs/code-practice-transparency-ai-generated-content
Commission FAQ: the code is voluntary, with machine-readable marking/detection and optional EU icons; reviewed at least every two years. - https://eur-lex.europa.eu/eli/reg/2024/1689/oj
Regulation (EU) 2024/1689, Article 50 — the legal basis for the transparency obligations the code supports.
Read next
Article 50: chatbots, deepfakes and the duty to make AI recognisable
The transparency obligations of Article 50 touch nearly everyone deploying AI towards the public: chatbots disclose that they are AI, synthetic content is marked machine-readably, deepfakes are labelled. The accompanying marking code of practice was published in final form on 10 June 2026.
AI in media and journalism: transparency, deepfakes and media freedom
AI touches journalism on three fronts: Article 50 of the AI Act requires transparency on deepfakes and AI text, the DSA regulates the spread of disinformation on platforms, and the European Media Freedom Act protects editorial independence and sources.
India's binding rules for synthetic content — a transparency regime without an AI law
On 10 February 2026 India notified G.S.R. 120(E), amending its IT intermediary rules to create a binding regime for 'synthetically generated information'. Without an AI law, India now mandates labelling, embedded provenance metadata and platform verification via intermediary liability.