Insight
EU AI Act: the transparency rules that apply from August 2026
Since 2 August 2026, Article 50 of the AI Act requires chatbots to disclose that they are AI and AI-generated content to be marked. What it means for your company.
Regulation (EU) 2024/1689, the AI Act, is being applied in stages. On 2 August 2026 the transparency obligations of Article 50 became enforceable. They apply to every organisation that offers or uses AI systems that interact with people or generate content in the EU, not only to large technology companies.
The four obligations
- AI interaction: systems that talk directly with people, such as chat assistants, must inform users that they are interacting with AI, unless this is obvious.
- Marking synthetic content: providers of generative AI must mark generated audio, images, video and text in a machine-readable way.
- Emotion recognition and biometric categorisation: deployers must inform the people exposed to such systems.
- Deepfakes and public interest text: deployers must disclose deepfakes and AI-generated text published to inform the public, unless a person has reviewed it and holds editorial responsibility.
What changed with the Digital Omnibus
The Digital Omnibus on AI, in force since 27 July 2026, postponed the requirements for high-risk AI systems listed in Annex III to December 2027. The transparency rules were not postponed. Generative AI systems that were already on the market before 2 August 2026 have until 2 December 2026 to implement machine-readable marking.
What we recommend
- List the AI systems your company provides or uses, including chatbots on your website and AI features in your products.
- Add a clear notice wherever people talk to an AI system.
- Check whether your AI tools can mark generated content, and how you label AI-assisted publications.
- Train your staff: AI literacy under Article 4 has applied since February 2025.
Fines for breaching the transparency rules can reach 15 million euro or 3% of worldwide annual turnover. Our own approach is described in our AI policy. If you need help classifying your AI use cases, contact our team.