EU agrees on framework of AI Act | ZFZ Postcard Cases

Last Friday, the European Union reached a provisional agreement on the adoption of the Artificial Intelligence Act (AI Act). With this achievement, the EU has the potential to become a pioneer as the AI Act will be the first comprehensive statutory framework that regulates artificial intelligence systems like ChatGPT or Llama.

The AI Act generally follows a so-called “risk-based” approach. The higher the potential risks of an AI system are the stricter the applicable rules are. Thus, the requirements that need to be complied with depend on the respective risk-classification. AI systems that are used in human resource management are classified high risk as they can have a significant impact on employees’ future career prospects. Violations can lead to fines ranging from EUR 7.5 million to EUR 35 million or from 1.5% to 7% of the global turnover.

In a next step, the details of the AI Act will be finalized. It is to be expected that the final proposal of the AI Act will be presented in the coming weeks. As it currently stands, the AI Act will principally become applicable 2 years after its entry into force, except for the prohibitions which will already be applicable after 6 months and the rules on General Purpose AI which will apply after 12 months.

For further information on this topic, please reach out to Hans Georg Laimer and Melina Peer.

The official press release by the European Commission can be found here.