Truly Risk-Based Regulation of Artificial Intelligence. How to Implement the EU's AI Act, a lecture by Martin Ebers, Professor of Law, was held on 6 September at the John Lukacs Lounge. Professor Ebers, President of the Robotics & AI Law Society (RAILS) in Germany, Professor of IT Law at the University of Tartu, one of the most renowned experts on European AI regulation, founder and editor-in-chief of the journal LegalTech, and founder and editor of the Cambridge Forum on AI: Law and Governance, was the guest of the Institute of the Information Society.
The expert gave a presentation on the key issues of the recently adopted European Union (EU) Artificial Intelligence Act (EU AI Act) and the risk-based approach it represents. Professor Ebers sought answers to questions such as: what does this approach mean; does the AI Act really take such an approach; what are the practical implications of compliance with the law; and how will it be enforceable in the case of GPAIMs (generative AI).
The regulation claims to be based on a risk-based approach - in order to avoid excessive regulation and respect the principle of legislative proportionality. But this is not necessarily the only correct approach to regulating AI, he warned. All the more so because, contrary to the claims of the European Commission and co-legislators, important provisions of the AI Act do not follow a truly risk-based approach, and Martin Ebers says there is ample evidence of this.
The presentation also covered the Professor's recommendations and the expected international impact of the legislation.
Text: Lilla Kovács
Photo: Dénes Szilágyi