Answer 9 questions. Get a personalised Article 50 readiness score. See where you stand, and what to do next.
Free · 3 minutes · No sign-up · results by email · 100% EU hosted
Article 50 of the EU AI Act (Regulation (EU) 2024/1689) sets transparency obligations for providers and deployers of certain AI systems. There is no exemption for SMEs or start-ups. Which duty applies depends on your role: the provider builds the system or puts it into service under its own name; the deployer uses it.
The legislation ensures citizens know when they are interacting with an AI (such as chatbots) or consuming synthetic content (text, image, audio, or video generated by AI).
| Obligation | Article | Who is responsible |
|---|---|---|
| Design AI systems that interact with people so those people are informed they are dealing with AI, unless this is obvious from the context | Art. 50(1) | Provider |
| Mark AI-generated audio, image, video and text outputs in a machine-readable format, detectable as artificially generated | Art. 50(2) | Provider of the generative AI system |
| Inform people exposed to an emotion recognition or biometric categorisation system that it is operating | Art. 50(3) | Deployer |
| Disclose that deepfake image, audio or video content was artificially generated or manipulated | Art. 50(4), first subparagraph | Deployer |
| Disclose AI-generated or manipulated text published to inform the public on matters of public interest, unless it went through human review or editorial control under someone's editorial responsibility | Art. 50(4), second subparagraph | Deployer |
| Give this information clearly and distinguishably, at the latest at the first interaction or exposure, meeting accessibility requirements | Art. 50(5) | Provider and deployer |
An SME that builds a chatbot on a third-party model and offers it under its own name may be the provider of that chatbot. Role assessment can require specialist review.
Providers must design AI systems that talk to people so those people are told they are interacting with AI, unless this is obvious to a reasonably well-informed person in the context. If your company built or branded the chatbot, treat this duty as yours.
Providers of systems that generate synthetic audio, image, video or text must mark outputs in a machine-readable format so they can be detected as AI-generated. Assistive editing that does not substantially alter the input is exempt. Generative systems placed on the market before 2 August 2026 have until 2 December 2026.
Deployers must inform the people exposed to these systems that they are operating, and process personal data in line with the GDPR. Emotion recognition in the workplace or in education is prohibited by Article 5, except for medical or safety reasons.
Deployers must disclose that deepfake image, audio or video was artificially generated or manipulated; for evidently artistic, creative, satirical or fictional work, a lighter disclosure applies. AI-generated text published to inform the public on matters of public interest must be disclosed, unless a person reviewed it and holds editorial responsibility.
A deepfake is AI-generated or manipulated image, audio or video that resembles existing persons, objects, places, entities or events and would falsely appear authentic (Art. 3(60)).
Take the free benchmark, then check your own AI notices and terms with Themio at launch.
Last reviewed: 20 October 2026. General information, not legal advice. Sources: Regulation (EU) 2024/1689, Articles 3(60), 5, 50 and 99; Regulation (EU) 2026/1744; Code of Practice on Transparency of AI-Generated Content (10 June 2026).