AI Act greenlighted by the Council of the EU

IRIS 2024-6:1/3

Justine Radel-Cormann

European Audiovisual Observatory

Following our recent article on the adoption by the European Parliament of the Artificial Intelligence Act (AI Act) in March 2024 (2024-3:1/3), the Council of the European Union greenlighted the text on 21 May 2024. 

Article 50(4) imposes transparency obligations on deployers of AI systems when they generate or manipulate content (deep fake): the content shall be labelled as generated/manipulated. An exemption to the rule applies when said content forms part of an evidently artistic, creative, satirical or fictional work: the transparency obligation is then limited to the disclosure of the existence of the generation/manipulation in an appropriate manner that does not hamper the display or enjoyment of the content. 

The transparency obligation applies to generated/manipulated content which is published for the purpose of informing the public on matters of public interest. An exception to this obligation takes place when the AI-generated content has undergone a process of human review or editorial control and where a natural or legal person holds editorial responsibility for the publication of the content. 

Article 53(1)(c) and (d) require providers of general-purpose AI models to comply with the reservation of rights as established by Article 4(3) of the Copyright Directive 2019/790. The providers shall make publicly available detailed summaries of the input training the machine, according to a template the AI Office is expected to provide in the near future. 

At the time of writing, the AI Act has not yet been published in the European Union's Official Journal. Once it is published in this official journal, it will enter into force 20 days later.