Original broadcast date: 3 Nov. 2023

The alignment problem in artificial intelligence raises important questions about ensuring that AI foundational models and systems align with human values. This panel examined how evolving safety research can be effectively integrated into AI governance and looked at the current state of play of the EU Artificial Intelligence Act with regard to foundational models. The panel discussed how codes of conducts, standards and oversight can support alignment. What can we learn from other relevant frameworks such as the EU General Data Protection Regulation or the novel systemic risk mitigation and co-regulatory approach in the EU Digital Services Act.

Panel: Leonardo Cervera-Navas, Dylan Hadfield-Menell, Cornelia Kutterer, Philippe Verreault-Julien.