All guides

Article 50: AI disclosure and content labelling rules, explained

Chatbot disclosure, machine-readable marking, deepfake labels. The transparency rules enforceable since 2 August 2026, and the narrow December transition.

Published by Kritia · legal snapshot reviewed 1 September 2026

The four duties

1. Providers must inform people when they interact directly with an AI system (unless obvious from context). Your website chatbot must identify itself as AI.

2. Providers of generative systems must mark AI-generated audio, image, video and text output in machine-readable format where technically feasible.

3. Deployers of emotion-recognition or biometric-categorisation systems must inform people exposed to them.

4. Deployers must visibly label deepfakes, and AI-generated text on matters of public interest published without human editorial review.

The December 2026 transition: narrower than you think

The Digital Omnibus gave providers of generative systems already on the market before 2 August 2026 until 2 December 2026 to implement machine-readable marking. That's the only transition. It does not delay chatbot disclosure, deepfake labels, or anything for newly launched systems.

Content generated before 2 August 2026 does not need retroactive labelling.

Quick compliance moves

Rename your chat widget to 'AI assistant' and disclose in the first message. Turn on content credentials (C2PA) where your tools support them. Add visible AI labels to unreviewed published AI text. Document all of it.

Primary sources

The legal snapshot for this guide was checked against the documents below on 1 September 2026. Links go to the institution that published each source, not to a summary of it.

  1. Regulation (EU) 2024/1689 — Artificial Intelligence ActOfficial Journal of the European Union
  2. Regulation (EU) 2026/1744 — Digital Omnibus on AIOfficial Journal of the European Union
  3. Guidelines on Article 50 transparency obligationsEuropean Commission
  4. Transparency obligations under Article 50 — questions and answersEuropean Commission / AI Office

When this guide goes out of date, you’ll hear it here

This page was last reviewed on 1 September 2026. The wording behind it has already moved once this year. Kritia Law Watch is a free daily email for firms that use AI tools rather than build them: one thing that matters, published decisions kept separate from proposals, every claim linked to the authority that published it. A minute to read.

Enter your email; no account or card is needed. Double opt-in, one-click unsubscribe, never shared. Every issue is published in the open — read today’s first.

Compliance information, not legal advice. Basis: Regulation (EU) 2024/1689 as amended by Regulation (EU) 2026/1744, directly applicable in all EU member states; national implementing laws vary by country. Last reviewed 1 September 2026.