The grand event that was Mark Carney’s investment summit was preceded by bigger news, it being revelations about the human extinction threat posed by advances in artificial intelligence.
We’d heard horror stories before, but an Anthropic employee took fears to a new level by saying there was a 10-per-cent chance AI could kill all humans within the next decade.
To give the doomsday scenarios more gravitas, there followed the remarkable sight of the titans of the AI industry coming forward to confirm them. Runaway AI has to be brought under control, they said.
But the ominous soundings, which have seen the debate over AI regulation exploding in the U.S., were not enough to put a damper on the Toronto investment party. Reaction from the Carney government to the AI uproar was rather muted. It’s always been rather muted about AI’s dangers because this is a domain in which Canada has a competitive advantage. Ottawa wants to ensure we capitalize on the AI boom. Why highlight the hair-raising stuff? Why push for heavy regulation of the great growth engine?
But that won’t be an easy play to sustain now. The Armageddon scare has cast a pall, tilting the market from enthusiasm to trepidation. Canadian public opinion is likely to exert strong pressure on the Liberals to get off their derrières.
Ottawa had regulatory legislation all set – the Artificial Intelligence and Data Act – but it died on the order paper in the last parliament. Now in its place is a patchwork quilt of measures weaker than what that legislation would have provided.
Dario Amodei, Anthropic co-founder and CEO, and Marc Benioff, Salesforce CEO and co-founder, attend Dreamforce 2026 summit in San Francisco, on Tuesday.Carlos Barria/Reuters
They look especially anemic when we hear Anthropic chief executive Dario Amodei sounding alarm bells about how AI agents could effectively take control of large portions of the internet through co-ordinated agent swarms. And how they could become capable off devastating cyber warfare. And how advanced AI could make nightmarish biological weapons easy for rogue actors to create.
On the military front we’re already at the stage where AI-powered combat drones, as seen recently in an attack on Ukraine, can autonomously – i.e. with no supervision from earthlings – choose targets and strike them.
Responding to the warnings of the AI bosses, Mr. Carney told Bloomberg this week there’s a need for a more co-ordinated international response. “We’ve had initial discussions at the G7 of how that might be done,” he said. “Ultimately, a technology stability board, in our view, along the lines of a financial stability board, would make sense.”
But a board is hardly enough to quell fears.
“Canada has binding legal regulations on everything from guns to butter,” NDP Leader Avi Lewis posted on X Monday. “And none for AI.” It’s time for our government to stop “cheerleading and start regulating in the public interest,” he said.
Complicating the situation is Donald Trump’s stance. Even though the people making the technology were the ones making the dire warnings, the “very stable genius” (as he has called himself) was dismissive. There should be no slowing of AI development, he said; the fears being raised are a “hoax.”
Despite existential threats, the U.S. President is full steam ahead. He regards the climate crisis as a hoax too. He tears down programs to address it. He was once opposed to forever wars. Not so, given his take on Ukraine and Iran, any more. He still likes tyrants, a sampling being his high-five rapport with North Korean head case Kim Jong Un.
On AI regulation, Ottawa is limited in what it can do given the dominance of Americans in the field. But some restraints, like requiring companies to meet regulatory safety standards before their AI systems are deployed, are possible.
Questions are growing about whether artificial intelligence is outpacing safety measures for the technology. Reuters Technology Correspondent Deepa Seetharaman explains what's driving the sudden push to 'slow down' AI development, and weighs in on whether slowing the race is even possible.
Reuters
Mr. Trump’s basis for green-lighting AI is the fear, legitimate enough, that the U.S. will lose the AI race to China. To avoid that, Mr. Amodei says that like in the Cold War days when there were arms control negotiations on nuclear weapons, there could now be AI arms control negotiations between the U.S. and China.
Mr. Carney has credibility among middle powers. He could help build a coalition to pressure for AI safety and megapower co-operation on its emergence. “Middle powers must act together,” Mr. Carney said at Davos, “because if we’re not at the table, we’re on the menu.” Menu? With the lead the U.S. is getting on AI, competitors will be all the more vulnerable to subjugation by leaders like Mr. Trump.
Ottawa has been passive up to now on the threats of runaway AI without getting much pushback. With the new revelations on the catastrophic risks, there can be no passivity now.