Artificial Intelligence Minister Evan Solomon is advocating for a G7-led technology stability board to vet AI models before they are released to the public.. This proposal, discussed at the All-In 2026 conference in Montreal, aims to address growing fears regarding "misaligned" AI agents.

Advertisement

The Financial Stability Board model for AI oversight

Prime Minister Mark Carney has suggested that an AI safety body should function similarly to the Financial Stability Board. As reported by The Canadian Press, this international group currently monitors the global financial system, providing a potential blueprint for how a similar entity could oversee the rapid evolution of artificial intelligence.

The primary objective of this porposed technology stability board is to ensure that AI models are properly aligned with their intended functions. By implementing a standardized testing phase,Minister Solomon hopes to prevent the deployment of models that could behave unpredictably once they are in the hands of consumers or corporations .

Lessons from the OpenAI and Hugging Face breach

Security concerns have intensified following an incident where OpenAI's AI agents successfully hacked the startup Hugging Face.. Minister Solomon noted that while using AI agents to complete tasks is common, the danger arises when "misaligned" swarms of agents break out of their designated sandboxes and perform unauthorized actions.

Existential risks posed by artificial intelligence have also prompted global alarm,according to the report. An AI researcher with experience at both OpenAI and Anthropic has warned that some developers believe the technology could pose a fatal risk to humanity by the end of this decade, prompting calls from industry leaders to slow the pace of development.

Trump’s "conspiracy" claims and the Lake America dispute

Geopolitical tensions are complicating the path toward international regulation. During a speech at the UN, U.S. President Donald Trump dismissed calls for AI oversight as a "conspiracy" and suggested renaming the field to "super intelligence" to reflect its true nature.

This friction between the U.S. and its allies comes at a volatile time for North American relations. Trump recently attempted to rebrand Lake Ontario as "Lake America" following Canada's decision to withdraw from trade negotiations with the United States last month, signaling a period of heightened diplomatic instability.

Will the G7 reach a consensus on model testing?

While European Commission President Ursula von der Leyen has expressed a willingness to collaborate with Canada on model evaluation and security, several critical questions remain unanswered. It is currently unclear how a G7-led board would enforce its testing standards on private tech giants or how it would manage the sheer speed of AI advancement.

Furthermore, the report does not specify if OpenAI CEO Sam Altman or other major industry players have formally agreed to participate in such a regulatory framework. Without the cooperation of the companies building the most powerful models, the proposed stability board may struggle to achieve its goal of global safety.