Regulation would encourage AI innovation, uOttawa prof argues after Paris summit

Regulation of artificial intelligence would encourage — not stifle — innovation, says a University of Ottawa professor who presented at last week’s global AI summit.

Florian Martin-Bariteau, who is also the university’s chair of technology and society, flew across the Atlantic to address world leaders in Paris at a summit meant to set a global agenda for the rapidly developing technology.

Such international summits have been held annually since 2023 and bring together world leaders, researchers, and AI companies.

“Not any kind of innovation is good for society,” Martin-Bariteau said in an interview with CBC’s Ottawa Morning.

“Regulation is not always bad. It can help provide a playbook for innovators, know what they can do, not do, and innovate in a safe environment.”

Martin-Bariteau said he was hopeful the summit would allow leaders to develop concrete plans to handle AI, but instead faced a reality that was like “a cold shower.”

Conversations at the summit were “light toward commitment” and “lacking big actions,” he said. “A lot of people think regulation will stifle innovation.”

That was certainly the tone set by American Vice-President J.D. Vance, who told the summit “excessive regulation of the AI sector could kill a transformative industry just as it’s taking off.”

Despite the uphill battle, Martin-Bariteau said it’s necessary for tech experts to keep pushing for regulation, as artificial intelligence is already being used to interfere in democracies around the world.

Not ‘if’ but ‘when’

Martin-Bariteau, alongside University of Montreal professor Catherine Régis, led a team of international experts studying how AI is being used to affect elections and the democratic process.

Among other examples, the team studied a case in Romania, where the first round of the country’s presidential election was annulled in December because of a Russian campaign to promote one candidate, Calin Georgescu, on social media.

Martin-Bariteau’s policy brief, which was provided to attendees at the summit, describes how Russia used AI to create accounts and pro-Russian content that could be spread through a large network on social media.

“It’s not a question of if” AI will interfere in global democracies, Martin-Bariteau said. “It almost certainly will happen. How, and to what extent, is the big question mark.”

Barry Sookman, a Toronto-based lawyer with expertise in tech and AI, agreed. Sookman argued AI could ultimately make democratic institutions less trusted and less capable.

“That’s what our foes would do, undermine credibility of the governments, spew misinformation about government policies and the effects, and try to divide [us],” he said. “That’s a real danger.”

Martin-Bariteau said he’d like to see Canadian political parties adopt a code of conduct on the use of AI — which some have done in Europe but none have done in Canada.

“In 2025, we have at least two elections, the [Ontario] provincial and the federal one,” he said. “We need to get ready.”

Over-regulation concerns

Sookman said he understands concerns about over-regulating AI, if Canada wants to be “at the top of the innovation race.”

“If the countries that we do significant trading with have opted not to have a heavy hand in regulation, I think that we ought to be very, very cautious before we move ahead with regulation that stifles innovation in Canada.”

Sookman was one of many experts who spoke out against Canada’s Artificial Intelligence and Data Act (AIDA) when it was first unveiled, arguing stakeholders were not consulted and the act amounted to “nothing but a shell of a law.”

He said “targeted regulation,” however, could deal with serious problems like misinformation and AI interference in elections without undermining innovation and at the same time building trust and confidence in the technology.

Martin-Bariteau said he and other technology experts have heard worries about stifled innovation for years.

Stifling innovation was a concern expressed by businesses when Canada’s voluntary AI code of conduct was released, but signatories of the document later told the Canadian Press they didn’t regret signing it and wished more businesses would join them.

The same arguments were made about the internet and social media when those technologies were new, Martin-Bariteau said.

“We know them and we listen to them with kind of a smile,” he said.