When Article 50 transparency obligations apply, who must inform users, and what changed after the Omnibus.
Upcoming
Users must be informed when interacting with an AI system; AI-generated or manipulated content, including deepfakes, must be disclosed; disclosure duties apply for emotion-recognition and biometric-categorization systems. Article 49 EU database registration framework and governance/enforcement provisions also apply. This date was not deferred.
Changed by Omnibus Original: August 2, 2026
The machine-readable marking/watermarking duty under Article 50(2) applies to generative systems placed on the market before Aug 2, 2026 after a four-month grace period. Systems placed on the market on or after Aug 2, 2026 must comply from placement.
Changed by the Digital Omnibus on AI, pending Official Journal publication for legal effect.
The main Article 50 transparency milestone is rendered above from the deadline data. It covers direct AI interaction notices, AI-generated or manipulated content disclosures, deepfake disclosures, and disclosures for emotion-recognition and biometric-categorization systems.
No. The Omnibus changed the machine-readable marking grace period for some older generative systems, but the main transparency date was not deferred.
Some duties are provider-facing and some are deployer-facing. A deployer using AI to interact with people, disclose deepfakes, or run emotion-recognition or biometric-categorization systems should treat this page as a scoping prompt, not final legal advice.