AI Act: chatbots must say they’re AI, deepfakes must be labelled

Since 2 August the EU enforces the AI Act’s transparency rules; a ban on ‘nudifier’ apps follows on 2 December.

Since 2 August 2026 the Commission’s AI Office and national authorities enforce the AI Act. Chatbots and other interactive AI must tell you that you are dealing with AI, not a human, and AI-generated or edited images, video and audio must be labelled as such. In June 2026 Parliament and Council agreed changes to the Act: AI systems that generate child sexual abuse material or intimate images without consent (‘nudifier’ apps) are banned from 2 December 2026, while the rules for high-risk AI — such as credit scoring and hiring — were pushed back to December 2027.

For youYou have a right to know when you are talking to a bot. Be sceptical of unlabelled images and video that look real.

Next: 2 December 2026: nudifier ban, and machine-readable marking becomes compulsory for AI systems already on the market before 2 August. 2 December 2027: rules for stand-alone high-risk AI. 2 August 2028: high-risk AI built into products.

Verified against the sources above on 29 September 2026. Plain-language summary — not the legal text and not legal advice.