Art. 50

Article 50: Transparency Obligations

Article 50 requires disclosure that a person is interacting with an AI system, machine-readable marking of AI-generated synthetic content, information to people exposed to emotion recognition or biometric categorisation systems, and labelling of deepfakes and AI-generated text on matters of public interest. These duties have applied since 2 August 2026 and were not postponed as a whole by the Digital Omnibus; only Article 50(2) got a limited transition arrangement for pre-existing systems.

providerdeployer limited-risk
When it applies
What it requires
What you keep on file
Commonly misread

Article 50 as a whole is frequently reported as "delayed to December 2026" because of the Article 50(2) transition; in fact only the synthetic-content marking duty for pre-existing systems moved, new systems and all other Article 50 paragraphs have been binding since 2 August 2026 without a grace period.

Unsettled: The legal status of the Code of Practice on Transparency of AI-Generated Content, confirmed by the Commission on 20 July 2026, and whether it carries a presumption-of-conformity effect similar to the GPAI Code of Practice, was not conclusively established in the underlying research.

Recent activity

What regulators published on Art. 50

  1. Loading from the public feed…

Informational content only. Not legal advice and not a substitute for qualified counsel. Dates reflect Regulation (EU) 2026/1744 as of 04 Aug 2026.