UK, Canada forge AI safety partnership amid defence cooperation talks

Andy Burnham and Mark Carney sat down in Liverpool this week for their first formal bilateral meeting as prime ministers of Britain and Canada, a conversation that stretched across two urgent domains: how their governments should manage the risks of artificial intelligence, and how their militaries might work more closely together. The meeting concluded with both leaders watching Everton play football at their stadium—a moment of ease after substantive talks on matters neither nation has yet figured out how to regulate.

The stated outcome was straightforward enough: governments and technology companies need to work in partnership to protect people online and manage the dangers that AI systems pose. A statement from Burnham's office framed this as obvious common sense, though the fact that it required a bilateral meeting to say so hints at how unsettled the terrain remains. Days before the Liverpool gathering, Burnham's spokesperson had welcomed what he called a positive shift in the industry itself—major AI developers were finally talking openly about risks alongside opportunities. But the spokesperson also sounded a note of caution: policy decisions had to rest on evidence, not guesswork, and the existing rules that govern technology in both countries might not hold up as AI systems grow more powerful.

Neither Britain nor Canada currently has a law written specifically to govern artificial intelligence. Britain chose a different path in March 2023, releasing a policy paper that embraced what officials call a "pro-innovation" approach. Rather than pass a standalone law, the government handed responsibility to existing sector regulators—the financial watchdog oversees AI in banking, the health regulator watches over AI in medicine, and so on. It is a distributed model, betting that industries know their own risks best. Canada attempted something more direct. The government drafted the Artificial Intelligence and Data Act, a comprehensive measure designed to set rules across the economy. But in January 2025, parliament was prorogued—suspended by the Prime Minister—and that procedural move killed all pending legislation. The AI bill died with it and has not been revived. For now, Canada relies on privacy law and a voluntary code that industry players have agreed to follow.

The defence portion of the meeting touched on practical complications. Britain and Canada are trying to align their blueprints for a joint military financing bank that would serve allied nations. The two countries are also navigating Canada's application to join the UK-led Joint Expeditionary Force and its role as an observer in the Global Combat Air Programme. These are the mechanics of alliance-building, the unglamorous work of making sure two militaries can actually work together when called upon.

But the meeting also reflected a broader shift in Canada's geopolitical position. Ottawa is simultaneously pursuing closer ties with Europe, having applied to become the European Union's first "associate member." That move signals how much the trade tensions between Canada and the United States have unsettled the traditional North American alignment. European Commission President Ursula von der Leyen has backed the idea. For Britain and Canada, then, this bilateral meeting happens against a backdrop of Canada hedging its bets, drawing closer to Europe even as it deepens defence cooperation with its traditional ally across the Atlantic. The question of how those relationships will balance—and whether they can coexist without friction—remains open.