Silicon Valley tech executives and government leaders are ensnared in a debate over whether artificial intelligence could annihilate humankind, but Canada's AI and Innovation Minister Evan Solomon says the country is taking a more pragmatic and proactive path to address the risks the technology poses.
The likes of OpenAI LLC, Anthropic PBC and Google LLC have been calling for a slowdown in the pace of AI development after several researchers voiced concerns that AI "could kill us all by the end of the decade."
United States President Donald Trump, however, said such fears are a "hoax" and rejected any in development and efforts to regulate AI so the U.S. can maintain its lead over China.
Other tech industry voices have said those urging a slowdown amount to collusion by a few companies seeking to preserve their lead in AI.
But Solomon on Wednesday said there are more choices available between a slowdown and letting AI run rogue, pointing to Canada's "multi-pronged" approach of domestic policy and global collaboration.
"Canada is clear-eyed about the challenges, concerns and opportunities of AI. We take the risks seriously," he said. "But we're not passive. We're going to build (AI) actively. We're going to be the makers."
On that front, the federal government on Wednesday announced a joint $300-million investment with Germany in Montreal-based AI safety non-profit LawZero to develop an advanced and auditable AI system that can keep AI agents in check. Canada will contribute $150 million and Germany will give 100 million euros to the organization founded and led by Yoshua Bengio. LawZero will also set up an office in Berlin.
"This is a major investment (and) a serious effort to build safe, reliable Canadian AI," Solomon said.
He also said it complements the $50 million Ottawa earmarked for the Canadian AI Safety Institute under its national AI strategy released in June.
Earlier this week, Prime Minister Mark Carney floated the concept of establishing a global technology stability body akin to a financial stability board, which is an idea that has been "socialized," according to Solomon.
"We're working with our allies, I've talked to my counterparts about it and our national AI safety institutes are working together … to be more closely aligned on evaluation, safety and testing," he said. "We think that's one very positive way that you could regulate."
European Commission president Ursula von der Leyen on Wednesday said the European Union is eyeing more cooperation with Canada on AI model evaluation, verification and security.