What does AI transparency obligation mean?

AI transparency obligation is the umbrella term for statutory rules that are meant to ensure people can recognise when they are dealing with an AI system instead of a human – or when content is AI-generated. The idea behind it: the more convincingly AI systems imitate human communication, the more important it becomes to avoid deception.

In the EU AI Act, the transparency obligation relevant to chat and voice widgets is set out in Art. 50. It requires providers and deployers of AI systems that interact directly with natural persons to ensure that these persons are informed in a clear, distinguishable and accessible way that they are communicating with an AI.

Important for putting it in context: AI transparency obligation is not the same as the risk classification of AI systems that is also regulated in the AI Act (that is, the question of whether a system counts as "high-risk"). The only question here is whether users can recognise, at the moment of interaction, who or what they are communicating with – regardless of how "critical" the underlying AI application would otherwise be classified.

For website operators with a chat or voice widget, the transparency obligation under Art. 50 is the most practically relevant form of this principle. It applies from 2 August 2026 and potentially affects every company with an AI-powered widget on its own website, regardless of sector or size.

If you want to understand what this specifically means for your own website, you will find the full overview in the guide to the chatbot disclosure obligation.