AI Governance Club
🌍 AI Governance Club
AI agents are already producing software, conducting business activities, and automating personal tasks with only limited human oversight. The EU AI Act, written before this shift, was not designed for systems that plan, adapt, and act autonomously across long sequences of decisions. The result is a regulatory framework that applies to AI agents in principle but fails them in practice.
Join the AI Governance Club for a special author presentation of Session 26:
Regulating AI Agents by Kathrin Gardhouse, Amin Oueslati, and Noam Kolt
This is a rare opportunity to hear directly from one of the paper's authors. The analysis examines how the EU AI Act responds to five core governance challenges posed by AI agents: unreliable performance, misuse by malicious actors, privacy risks, inequitable access, and the breakdown of meaningful human oversight. The findings are consistent and sobering: policymakers in the EU and beyond will need to change course, and soon.
Presenter: Kathrin Gardhouse is a Senior AI Governance Associate at The Future Society and Policy Lead of AI Governance and Safety Canada. A lawyer and philosopher by training, she holds a PhD from McMaster University and advises policymakers on risks from general-purpose and agentic AI. She contributed to a legal commentary on the EU AI Act by the Institute for Law and AI and is certified by the IAPP as an AI Governance Professional.
For more information and to register for upcoming sessions: AI Governance Club