Shaping the Future of AI Regulation: Insights from the Quebec-Oxford-France Workshop in Montréal

As AI continues to transform societies and economies, the question of how to regulate it wisely and collaboratively is becoming one of the defining challenges of our time. On September 18 and 19, 2025, IVADO, the Université de Montréal, the Maison Française d’Oxford, the University of Oxford, the British Consulate-General in Montréal, the Quebec Government Office in London and the Canada CIFAR AI Chair held the second edition of the workshop serie “Shaping the Future of Artificial Intelligence Regulation” at the Court of Appeal of Quebec.

This international event, organized by Professors Catherine Régis (UdeM, IVADO), Célia Zolynski (Paris 1 Panthéon-Sorbonne), and Angeliki Kerasidou (Oxford), brought together about twenty invited participants from Quebec, the United Kingdom and France, including doctoral and postdoctoral researchers. Together, they explored how national and international approaches to AI governance can support innovation while protecting human rights, democratic values and public trust.

Crossed Perspectives on AI Regulation

The workshop opened with a rich discussion on how different jurisdictions are defining their regulatory paths. From the European Union’s ambitious AI Act to the United Kingdom’s flexible “pro-innovation” model and Canada’s still-evolving legislative framework, the exchanges compared the strengths and limitations of each model.

Célia Zolynski from University Paris 1 Panthéon Sorbonne used the example of deepfakes to demonstrate the need for truly interdisciplinary frameworks that combine law, technology, and culture. Rebecca Williams from the University of Oxford described the UK’s current position as one of observation and adaptation, relying on existing law, particularly judicial oversight, to regulate AI uses. Pierre Larouche from the Université de Montréal examined how Canada’s model balances risk, competition, and innovation in a shifting geopolitical context.

Presentations by Melissa Hyesun Yoon from Hanyang University in South Korea and Juan David Gutiérrez Rodríguez from the University of the Andes in Colombia expanded the debate to Asian and Latin American perspectives, reminding participants that AI regulation is a global issue deeply rooted in local contexts.

When Law Meets Code

The second session, chaired by Angeliki Kerasidou from the University of Oxford, focused on the complex task of turning principles into practice. Benjamin Guedj from Inria and University College London highlighted the technical limits of regulation, noting that transparency, auditing and model robustness require constant dialogue between lawyers and scientists. Christian Gagné from Laval University and IVADO advocated for national AI development aligned with local values and languages. Alexei Grinbaum from CEA-Saclay presented the AIOLIA framework, illustrating the practical application of the AI Act’s ethical principles and the tensions between security, privacy and efficiency depending on context.

Finally, judges Benoît Moore and Simon Ruel of the Court of Appeal of Quebec, joined by Paul Gagnon and Misha Benjamin of the law firm BCF, who discussed the role of lawyers in supporting companies and innovators deploying AI at both the local and international levels, as well as the challenges of implementing legislation in this context.

Global Governance and Geopolitics

The second day broadened the discussion to the international level. Benjamin Prud’Homme from Mila analyzed the geopolitics of AI and advocated for strengthened multilateral cooperation. Isabella Wilkinson from Chatham House emphasized the issue of transparency in a context of fragmented global governance.

Catherine Régis, representing IVADO, the University de Montréal, and Mila, presented the creation of the United Nations International Scientific Panel on AI, a major step toward embedding independent scientific expertise at the heart of international decision-making, while explaining the resulting scientific and legal tensions.

Toward Responsible Governance

Throughout the workshop, exchanges highlighted a shared conviction. Effective AI regulation must be based on clear principles and adapted to technological realities. It must rely on universal human rights values while remaining flexible in the face of rapidly evolving technologies.

The IVADO delegation, composed of Gaëlle Foucault, Antoine Congost, Halima Bachir, Clare Mulrooney and Emma Kondrup, enriched these discussions with new perspectives, representing the next generation of research at the interface of law, technology and public policy.

This conference strengthened ties between Quebec, France and the United Kingdom, an academic alliance expected to play a central role in building trustworthy AI. IVADO is proud to support these efforts through its research, training and knowledge mobilization programs, which aim to link scientific reflection with concrete action in responsible governance.