Amid growing concerns about the rapid advancement of AI and the lack of effective oversight from dominant superpowers, AI specialists and Canadian policy advisors are advocating for a coalition of middle powers to establish safety standards, with Canada in a leading role.
- Middle powers can create an independent AI safety authority.
- Canada’s leadership is backed by top AI experts including Yoshua Bengio.
- US and China are seen as insufficiently committed to AI safety guardrails.
What happened
A private gathering of AI policy experts, advisors, and advocates took place in Ottawa where the second International AI Safety Report was discussed. The event, conducted under the Chatham House Rule, highlighted the gaps in risk management as AI capabilities continue to evolve rapidly.
Participants expressed a strong interest in forming a middle-power coalition apart from the US and China to establish safety frameworks for AI technology. Canadian AI pioneer Yoshua Bengio emphasized the dual opportunity and risk Canadian and allied countries face in AI development, stressing safety as a priority.
Why it matters
The current AI development landscape is dominated by the US and China, both engaged in an AI arms race perceived as lacking sufficient safety guardrails to prevent harm or misuse. Middle powers, including Canada, could fill this leadership vacuum by offering a balanced approach to regulation and oversight.
Experts note that the US government’s unpredictability and perceived unreliability have led frontier AI developers to seek leadership from middle powers. This coalition approach could drive the global agenda for AI safety standards, mitigating potential economic, ethical, and security risks.
What to watch next
Observers should monitor whether Canada and its allies formalize this coalition to influence global AI safety norms, as calls by figures like Prime Minister Mark Carney for collective action have yet to gain significant momentum.
Additionally, how the US and other global superpowers respond—whether by setting standards themselves or collaborating with middle powers—will shape the international governance landscape of AI technology in the coming years.