Minister says he’s working with G7 to set up AI safety organization

OTTAWA — Artificial Intelligence Minister Evan Solomon said he spoke with his international counterparts in the G7 on Tuesday about setting up an organization to tackle AI safety.

The “technology stability board” would evaluate AI models for testing before they are released to make sure they are “aligned” with what they’re supposed to do, he said.

Prime Minister Mark Carney said last week it would make sense to create an AI safety board along the lines of the Financial Stability Board, an international group which monitors and makes recommendations about the global financial system.

Such an institution for AI has not yet been launched or announced.

Warnings about the risks AI poses to humans have prompted global alarm in recent weeks.

After an AI researcher who worked at both OpenAI and Anthropic claimed those building the technology “earnestly believe that it could kill us all by the end of the decade,” the heads of some of the biggest AI companies in the world said the pace of AI development should be slowed down.

Concern has swirled around AI bots since OpenAI disclosed over the summer that its AI agents hacked AI startup Hugging Face.

Solomon was asked about AI swarms Tuesday. He said that “these are terms to describe the use of AI agents that are used to accomplish tasks,” which is not uncommon.

“The place that this has been very concerning is when AI agents, in numbers, break out of a sandbox and are what’s called misaligned. They do things that they are not asked to do,” Solomon said.

He said the government is aware of situations like the Hugging Face incident and he has spoken about them with international counterparts and with OpenAI CEO Sam Altman.

Solomon said he also convened a meeting with G20 ministers on Friday to talk about safety issues.

He said concerns about “frontier models slipping out of control” are real but not new.

“But where the models are and how fast they’re getting to this point is going very quickly. And so the public is concerned about that, as are we. So we take that incredibly seriously,” he said.

Building safe and reliable AI “will require global regulation in the sense of model testing and evaluation,” Solomon said. “We’re working with our partners to see where this lands.”

European Commission President Ursula von der Leyen has said Europe will work with countries including Canada on issues like model evaluation, early warnings and AI security.

U.S. President Donald Trump has claimed the mounting calls for regulation are the product of a conspiracy. He repeated his anti-regulation message in a speech at the UN Tuesday and said he would change the term from “artificial intelligence” to “super intelligence.”

Solomon declined to comment on that Trump statement Tuesday.

“Canadians are not enthusiastic at all about renaming things, so let me leave that aside. It’s the substance that I think Canadians care about,” he said.

Trump attempted to rebrand Lake Ontario as Lake America after Canada pulled out of trade talks with the United States last month.

This report by The Canadian Press was first published Sept. 22, 2026.

Anja Karadeglija, The Canadian Press