Prime Minister Mark Carney painted an ambitious picture at the Liberal convention this week: a future where artificial intelligence lifts up all Canadians, not just the wealthy and connected.
"Our goal is AI for all," Carney declared at the Vector artificial intelligence research institute in Toronto. "AI governed by Canadian values, accountable to Canadians and serving Canadians."
It's a compelling promise. But according to leading democratic theorists, Canada's political system may be fundamentally broken when it comes to delivering on such ambitious technological governance.
Democracy in Crisis Mode
Political theorist Hélène Landemore from Yale University and democratic innovator Peter MacLeod—who has spent two decades experimenting with alternatives to traditional politics—argue that the problem runs deeper than party platforms or policy proposals.
"For the past 30 or 40 years, all the trend lines are pointing in the wrong direction," MacLeod explained in a recent episode of The Globe's Machines Like Us podcast. Voter turnout continues declining. Trust in government keeps eroding. Political parties have hollowed out, their memberships collapsing across the Western world.
The real culprit, Landemore argues, isn't just the system—it's how politicians are selected in the first place.
"The selection mechanism for legislators is oligarchic. It's no surprise that when you send representatives of socio-economic elites to power, you end up with a system overly responsive to the affluent," Landemore noted, citing decades of political science research proving exactly this point.
A Radical Rethink: Democracy Without Politicians
Rather than tinkering around the edges—removing money from politics, ethics reforms, better transparency—Landemore proposes something far more radical: sortition. In plain language, randomly selecting ordinary citizens to serve in the legislature, the same way juries are chosen for courts.
"If you had a truly representative sample of the public in power, you'd end up with laws that actually serve the majority," she argues.
MacLeod has been testing this theory through citizens' assemblies across Canada and internationally. These aren't focus groups or surveys—they're deliberative bodies where everyday people dive deep into complex policy questions, hear expert testimony, debate with peers, and reach informed consensus.
The Missing Piece: Public Participation
Both experts emphasize that the current system treats citizens as passive spectators rather than active participants.
"Most people are, at best, spectators to democracy," MacLeod explains. "We have a tendency to tell stories about a public that's too emotional, too ignorant, too apathetic to get involved. But I think apathy is just a fancy word for blaming the victim."
The real issue? People recognize they're not genuinely welcome in the political process, and the opportunities to make a real difference are vanishingly few.
What This Means for AI Governance
As Canada navigates the complex challenge of regulating artificial intelligence—a technology that touches everything from healthcare to employment to national security—the question becomes urgent: Can a political system that has lost public trust, that responds primarily to elites, and that excludes ordinary citizens from meaningful participation actually govern AI in everyone's interest?
Carney's vision of "AI for all" is admirable. But Landemore and MacLeod suggest that without fundamental democratic reform, the promise will ring hollow—and Canadians may find themselves locked out of decisions that will shape their future.
This article is based on reporting from The Globe and Mail's Machines Like Us podcast, hosted by Taylor Owen, featuring interviews with Hélène Landemore and Peter MacLeod. For the full interview and more analysis, visit The Globe and Mail.