The European Commission has published in the Official Journal of the EU guidelines for the application of transparency obligations for certain artificial intelligence systems under Article 50 of Regulation (EU) 2024/1689, known as the Artificial Intelligence Act. As stated in the document, the guidelines are intended for competent authorities, as well as for providers and users of AI systems, to ensure consistent and proportionate application of the rules.
The transparency obligations apply two years after the Act enters into force, that is, from 2 August 2026. The document elaborates four types of obligations: informing users that they are communicating with an AI system, marking synthetic content in a machine-readable format, notifying persons exposed to emotion recognition or biometric categorisation systems, and marking deep fake content and text published to inform the public on matters of public interest.
The guidelines also provide for exceptions, among others when the artificial nature of the interaction is obvious, when the system performs an auxiliary function of standard editing, or when the text has undergone human review and editorial responsibility. A special disclosure regime is envisaged for deep fake content in artistic, creative, satirical or fictional works. The Commission emphasises that the guidelines are not binding and that only the Court of Justice of the EU can give the final interpretation of the Act.
According to the document, the information from Article 50 must be provided in a clear and distinguishable manner at the latest at the first interaction or exposure, while respecting accessibility requirements.
What this means for businesses
For companies that use or develop AI systems, the guidelines provide a practical framework for compliance before the August 2026 deadline. AI system providers should incorporate mechanisms for marking synthetic content and informing users, while users of emotion recognition systems and deep fake tools must plan disclosure obligations. Since the guidelines are not binding, legal certainty will depend on the interpretation of the Court of Justice of the EU and the practice of supervisory authorities. The risk of sanctions and reputational damage is higher for companies that publish content without a clear label of origin.
This summary was written with AI based on the cited source and checked by automated rules. For exact details, read the original article.