Canada has joined a France-led international declaration on artificial intelligence safety that calls for governments and companies to share information and develop early-warning systems for serious cyber and AI-control incidents, Global News reports.

The voluntary agreement brings together national governments and major AI developers, but its non-binding nature has prompted questions about how its commitments will be enforced.

According to reporting by Global News’ Alex Boutilier, initial endorsers include Japan, Norway, the Netherlands and Bahrain, alongside companies including Anthropic, OpenAI, Microsoft and Canada’s Cohere.

The declaration outlines three broad priorities: a common approach to managing risks from advanced AI, responsibility and accountability across the sector, and greater information sharing about major threats.

Signatories propose developing a public repository of non-sensitive AI risk assessments, using a standardized format that would make it easier to compare models. The declaration also calls for legal frameworks that provide certainty for developers and protect citizens in cases of large-scale damage.

“Frontier models transcend national borders in their development, deployment and impact, and their opportunities and risks are therefore inherently global,” says the statement, released by the Prime Minister’s Office on Monday.

Florian Martin-Bariteau, director of the University of Ottawa’s Centre for Law, Technology and Society, told Global News that such statements can signal the federal government’s approach but cannot replace regulation.

“I have nothing against anything in the statement, this is what we need … but if there is no teeth, if there are no people to do the follow-ups, the regulators, it will never work,” he said.

The commitment comes as Prime Minister Mark Carney’s government promotes AI as a source of economic growth and innovation. Global News reported that Carney announced a $1.6-billion federal investment in Cohere ahead of last week’s budget.

Sources