European Parliament Approves AI Act

On March 13, 2024, the European Parliament approved the AI Act (Artificial Intelligence Act, “Regulation“), the first general law on Artificial Intelligence.

The body of the Regulation consists of more than 450 pages, 180 recitals, 113 articles and 13 annexes.

The Regulation aims at regulating the development, provision and use of AI systems in Europe. It will apply not only to European AI system providers, but also to non-European ones in case the outputs produced by AI systems are used in Europe. Therefore, the Regulation will be applicable to all companies and public entities using AI systems in the territory of the European Union or using the output of the system in the European Union. It will also be applicable to all entities categorized as importers or distributors. The Regulation provides for several exemptions (e.g. military purposes, national security etc).

Governance will be on a dual level: European and national.

At the European level, the European Artificial Intelligence Committee will be established, which will be composed of representatives from each member state. At the national level, on the other hand, each country will have to establish a National Supervisory Authority, which will be responsible for overseeing the implementation of the regulations and the imposition of sanctions. The penalties provided for in the Regulations include fines of up to 35 million euros or 7 percent of the previous year’s total annual worldwide turnover (whichever is higher).

The Regulation shall now be subject to a final check and is expected to be finally adopted before the end of the legislature.

The entry into force of the Regulation will be gradual, in fact it will enter into force 20 days after publication in the EU Official Journal and will begin to apply 24 months after entry into force except for:

  • bans on prohibited practices, which will apply from six months after entry into force;
  • the codes of good practices (9 months later);
  • rules on general purpose AI systems, including governance (12 months later);
  • obligations for high-risk systems (36 months later).