Regulation

EU AI rules begin implementation: mandatory disclosure for chatbots and labelling for deepfakes

Implementation of the EU's artificial intelligence rules has begun, introducing mandatory disclosure when users interact with chatbots or other interactive AI systems and requiring labelling for deepfakes.

EU AI rules begin implementation: mandatory disclosure for chatbots and labelling for deepfakes

The European Union's new rules on artificial intelligence have started to be implemented. The measures introduce mandatory labelling requirements for interactive AI systems and for content that has been created or altered using AI. The stated aim is to inform users and reduce the risk of deception.

What the new rules require

  • Chatbots and other interactive artificial intelligence systems must clearly notify users that they are interacting with an AI system rather than a live person. This obligation applies to communication interfaces.
  • Deepfakes — images, videos or audio recordings edited or generated using AI — must be labelled so that consumers can identify their artificial origin.
  • AI-generated or AI-modified content must also include machine-readable markers. These markers are intended to facilitate automatic detection and handling of such content.

Why these measures matter

According to the European Commission's Representation in Hungary (Európai Bizottság Magyarországi Képviselete), the rules aim to both reduce deception and manipulation and to help people make better-informed decisions. Human-readable labelling together with machine-readable signals are intended to make it easier for consumers and platforms to distinguish between genuine human communication and AI-produced content.

Business obligations and proving compliance

The Representation says the rules clarify the responsibilities of businesses and provide practical means for demonstrating compliance. Machine-readable markers and mandatory disclosure in user-facing interfaces are examples of measures that can help companies document and verify conformity with the rules.

Practical implications

Implementation of the regulation is expected to require changes in how digital service providers operate: operators of chat interfaces, content creators and platforms will need to adopt technical and informational solutions to meet the obligations. The overarching goal is to give users clearer visibility into when they encounter content that has been generated or processed by artificial intelligence.