OTTAWA — With a suite of artificial intelligence regulation laws now before Parliament, AI Minister Evan Solomon says the government has the tools it needs to tackle the issues undermining public trust in the technology.

On Monday, Solomon tabled his long-awaited data protection bill to update Canada’s antiquated private-sector privacy laws, which includes stiff fines for firms that violate the rules. The introduction of that act capped a flurry of related moves: in the past six months, the government has put forward legislation to regulate social media and AI chatbot companies, ban the use of deepfakes in elections and make it a criminal offence to share non-consensual sexualized deepfakes. 

“I am extremely confident that, with this bill and the three other bills, we can do what we said: protect children, protect personal information and protect privacy,” Solomon said in an interview with The Logic. 

Earlier this month, Prime Minister Mark Carney acknowledged Canadians have a ways to go before they’re ready to embrace the new technology—not least coming to terms with its privacy implications. 

“Globally, Canada ranks near the bottom of countries in AI training in literacy and trust,” he said, noting that only 12 per cent of Canadian businesses use the technology. 

In a Leger poll last summer, a majority of Canadians surveyed expressed worry about their privacy; fears that society would become dependent on AI; and concern that the technology would threaten their jobs and spread false information during elections. 

Yet respondents also acknowledged the potential benefit of the technology: 60 per cent of those surveyed said it would improve efficiency and 44 per cent said it would reduce human error.

Solomon said the new laws, combined with a new regulator that has the power to enforce them, are a “very comprehensive way to gain trust” without putting an undue burden on companies. 

That equanimity could develop over time, said Florian Martin-Bariteau, a University of Ottawa law professor and director of the university’s AI + Society Initiative, but there is still much to do. “We have a good foundation, but can we build the house?” he said in an interview.

It will take at least 18 months for the government to create the new regulator after the digital safety bill makes it through the parliamentary process, Martin-Bariteau noted, and there will be no way to enforce the rules until that happens. “I think that the wild west is what is creating distrust toward AI and the digital economy,” he said. 

Martin-Bariteau said Ottawa has not yet explained how the government will regulate AI tools themselves. The previous Liberal government had tabled legislation aimed specifically at “high-impact” AI systems, and would have mandated new record-keeping, transparency and monitoring requirements for the products. That bill died when then-prime minister Justin Trudeau prorogued Parliament, and Solomon promised not to reintroduce it.

The previous government also encouraged companies to sign up to a voluntary code of conduct with similar, though non-binding, provisions that were in the now-defunct bill. Some 46 firms are currently listed as signatories to the code, including Cohere, Coveo, IBM and OpenText. 

Solomon said the voluntary system will remain in place as a way to encourage companies to foster a corporate culture of responsibility.