EU AI transparency rules take effect

New EU transparency rules for AI systems take effect on 2 August 2026, aiming to help people recognise AI-generated or manipulated content and understand when they are engaging with an AI system. The measures target risks including misinformation, manipulation at scale, fraud, impersonation, and consumer deception.

Certain AI-generated or altered images, audio, and video resembling existing people, objects, places, entities, or events must be clearly labelled and include machine-readable marks. The requirements also cover emotion recognition and biometric categorisation tools, as well as text published to inform the public on matters of public interest when there has been no human review or editorial control.

Users must be clearly informed when they are interacting with systems such as chatbots, AI agents, and avatars rather than a real person. The Commission has issued guidelines explaining how providers and deployers can demonstrate compliance, including through adherence to a code of practice.

National market surveillance authorities, the European AI Office, and the European Data Protection Supervisor will enforce the rules. Authorities may issue fines up to €15 million, or 3% of global annual turnover, for companies and up to €750k for EU institutions, bodies, and agencies, with proportionality considered for SMEs and SMCs. The AI Act entered into force on 1 August 2024 and applies in stages.