Carney government balancing AI regulation against the promise to unlock its potential

Hamilton Spectator

a day ago

OTTAWA—Prime Minister Mark Carney believes artificial intelligence is key to unlocking Canada's economic potential.
He carved out a specific ministry for the file, installing former broadcaster Evan Solomon as the country's first-ever minister of AI.
The 'transformative nature' of the technology garnered a mention in the sole mandate letter he issued for his cabinet.
Capitalizing on the use of AI is one of Carney's three 'core missions' as he prepares to host next week's G7 leaders' summit.
Yet among the nascent Liberal government's ambitious promises to return Canada to its position as a global AI leader, promote the widespread adoption of the technology, and invest in the infrastructure needed to do so, there are no concrete pledges to regulate the sprawling uses of the rapidly evolving tools.
'AI is a fundamentally transformative technology and has the capacity to change the way we do almost everything. So I see this as just a point in history where we are transforming the way our markets work, the way our societies work, and we want that to be good,' said Gillian Hadfield, a professor of AI alignment and governance at Johns Hopkins University, and former member of the Canadian AI Advisory Council.
'When I look around the world, I see governments that have not really figured out: 'What do we need to do in the legal and regulatory space to manage this transition well?''
On Tuesday, Solomon danced around the issue of regulation, noting the difficulties of spurring AI development while also ensuring the technology is deployed responsibly.
'It's easy for editorials to write: 'Just find the right balance. Don't be so unconstrained as the U.S. and China, who see any regulation as a constraint on security or innovation. But don't be too overly protective like Europe,'' Solomon said at a Canada 2020 conference in Ottawa.
'OK. Perfect. Easy. Throw the dart, blindfolded, after six beers.'
What Solomon did suggest is that the Carney government will distance itself from the previous Liberal government's appetite for imposing regulations.
'We are moving from our back foot of just warning and overindexing on warnings and regulation, to our front foot, to make sure that the Canadian economy and all Canadians benefit from … using this technology productively,' he said.
Carney has pushed for advancing AI in myriad ways, from pledging to build data centres, proposing tax credits that would incentivize businesses to adopt the technology, and using it to improve government efficiency.
But his government has not addressed what will happen to the Artificial Intelligence and Data Act (AIDA), which was folded into a larger bill aimed at updating privacy laws and creating a regulatory framework for AI. The bill never became law due to the suspension of Parliament earlier this year and the triggering of a snap spring election.
The act was a particularly contentious prong of the proposed legislation, with critics blasting the act for concentrating too much power in unwritten regulations, concerns that having the same ministry simultaneously regulate and champion AI would introduce conflicts, and a lack of clarity on what AI systems it would apply to and what kinds of harms the legislation would minimize.
But some experts urged the government to swiftly pass the legislation, warning that Canada was falling 'out of sync' with the uncontrollable pace at which technology was being used.
Solomon referenced the bill Tuesday, saying he wouldn't 'abandon regulation,' but that Canada will need to 're-examine, in this new environment, where we're going to be on that.'
Ignacio Cofone, a professor of AI regulation at the University of Oxford and former Canada research chair in data governance at McGill University, told the Star in an email that it was critical that Canada move forward with an improved version of the act.
'AI systems already shape decisions in consequential areas as diverse as housing, employment, health care, and criminal justice, often in opaque and unaccountable ways,' Cofone wrote, adding that industry, which has 'incentives to prioritize profit' should not be left to regulate itself.
Two former senior government officials with knowledge of the previous government's AI strategy told the Star, on the condition they not be named, that they believe the Carney government will take a more hands-off approach.
One source said they believed that Carney is likely wary of the 'political mess' the Trudeau government found itself in as it crafted its legislation.
'Every day Canada doesn't advance its own responsible use policies or regulatory frameworks, we are just going to be further and further behind,' the source said.
Another source said that while they agreed that Canada cannot 'overregulate' the industry, particularly in the face of unprecedented trade disputes with the U.S., 'it shouldn't be controversial to say that we need to enshrine rights for Canadians against some of these uses.'
Teresa Scassa, a Canada research chair in information law and policy at the University of Ottawa, said Canada appears to be stuck between Europe — which last year passed the world's first AI regulation law — and the U.S., which has dismantled efforts to address the risks of the technology.
Scassa said a Canadian AI regulation law could be leveraged in the country's trading relationship with the EU, potentially allowing Canadian companies to do business in Europe.
'On the other hand, we have a government south of the border that sees everything through a trade lens. And if Canada has strong AI regulation or even weak AI regulation, that could be seen as a trade irritant,' Scassa said.
But Hadfield said the government should not be looking at innovation and regulation as mutually exclusive objectives.
'Our economies are built on good, reliable, legal infrastructure. And if the economy is changing so rapidly with this very different technology … then we absolutely need to be thinking very hard about governance,' Hadfield said.