EU AI Act transparency rules now apply (2 August 2026)

Effective date: 2 August 2026. Last checked: 12 November 2026.

What changed

The transparency obligations in Article 50 of the EU AI Act began to apply on 2 August 2026. In plain terms, people must be told when they are talking to an AI system rather than a person, and certain AI-generated or manipulated content must be disclosed or marked.

The EU's "Digital Omnibus" amendments, approved in June 2026, delayed the rules for high-risk AI systems: standalone high-risk systems to 2 December 2027 and AI built into regulated products to 2 August 2028. They did not delay the main transparency duties. The one concession is narrow: generative AI systems already on the market before 2 August 2026 have until 2 December 2026 to add machine-readable marking to their output.

Who it affects

Any business whose website chat, WhatsApp assistant or voice agent talks to people in the EU, and anyone publishing AI-generated content to an EU audience. That includes UK businesses serving EU customers.

What to do

• Make sure every AI assistant says clearly, at the start, that it is AI.
• Give people an easy way to reach a person.
• Keep a short record of which AI tools you use, where and why.

Every assistant we build is clearly labelled as AI and can hand over to a person. This is general information, not legal advice.