Days after a phone call with Pope Leo XIV about the moral stakes of artificial intelligence, Canadian Prime Minister Mark Carney stood in Toronto on Thursday and announced precisely the kind of national framework the pontiff had demanded. The strategy, branded “AI for All,” commits more than $2.3 billion in spending over five years. It is Canada’s most ambitious attempt yet to position itself as a serious player in the global AI race.

But the document has a conspicuous gap. For all its talk of protecting Canadians, it offers few concrete safety mechanisms, no hard timelines for new regulation, and no clear enforcement architecture. The Pope warned governments that AI “demands” to be disarmed.

Carney’s strategy reads more like an invitation to invest.

The six pillars

The plan is organised around six pillars, first outlined in April’s spring economic update: protecting Canadians and safeguarding democracy, empowering Canadians, powering shared prosperity, building a sovereign AI foundation, scaling Canadian champions, and building trusted partnerships and global alliances.

Artificial Intelligence Minister Evan Solomon, Canada’s first cabinet minister with the title, said the strategy reflects what citizens want. “Canadians want safe, reliable, and sovereign AI,” Solomon said in a statement. “They want the best tools to build a prosperous future guided by our values.”

The jobs numbers are the headline grabbers. Ottawa is targeting up to 90,000 AI-related jobs and work opportunities for young Canadians by 2031, plus a further 250,000 positions created through AI adoption across the broader economy. The strategy also aims to lift business AI adoption from its current rate of roughly 12 per cent to 60 per cent by 2034.

Sovereignty over safety

The document’s strongest language is reserved for sovereignty, not safety. Canada currently relies heavily on foreign cloud infrastructure, and the strategy frames this as a vulnerability. It proposes a “build-partner-buy” approach: build key capabilities domestically where possible, partner with trusted allies, buy from the market when appropriate.

Concrete measures include plans for a “world-leading” supercomputer and the expansion of sovereign data centres capable of 100 megawatts to serve Canadian clients. Up to $1 billion will go towards public supercomputing infrastructure alone.

The sovereignty push builds on an alliance Carney’s government has already begun constructing. In February, Solomon and his German counterpart signed a joint AI declaration at the Munich Security Conference, launching a Sovereign Technology Alliance designed to reduce dependence on concentrated technology providers.

The strategy says Canada will expand that alliance further. The parallel with Europe’s own tech sovereignty push is hard to miss.

The papal factor

Carney and Pope Leo spoke by telephone on 29 May, days after the pontiff released Magnifica Humanitas, his first encyclical. The document, signed on the 135th anniversary of Rerum Novarum, names AI as this generation’s industrial revolution and argues that without enforceable limits it will deepen inequality, erode human agency, and concentrate power among a handful of firms.

According to both the Vatican and the Prime Minister’s Office, the two leaders discussed the imperative that AI must serve humanity, beginning with the protection of the individual. Carney reportedly expressed Canada’s desire to lead internationally on responsible AI development.

The timing was deliberate. The Vatican had enlisted Anthropic co-founder Christopher Olah to speak alongside cardinals at the encyclical’s launch, signalling that the Church sees AI governance as a conversation requiring technologists at the table. Carney’s subsequent strategy announcement positions Canada as the first G7 nation to respond directly to the Pope’s call.

Where the safety details aren’t

The strategy promises new consumer privacy legislation enshrining a right to privacy and safeguarding children’s information online. It also pledges to modernise safety laws. But it provides no timeline for either.

Earlier this year, Canadian Identity and Culture Minister Marc Miller said the government was considering banning AI chatbots for children under 16. That restriction does not appear in the strategy. Officials say it is under review and may be folded into separate online harms legislation expected later this year.

The omission is notable given the global context. The EU is building its own sovereign AI infrastructure while simultaneously enforcing the AI Act, the world’s most comprehensive regulatory framework. Canada’s previous attempt at AI legislation, the Artificial Intelligence and Data Act within Bill C-27, is widely regarded as inadequate and has not been revived.

Literacy and training

On the workforce side, the strategy introduces a national AI literacy initiative offering free entry-level training to all Canadians. Ottawa plans to reach one million post-secondary students and train more than 3,000 educators with AI learning kits.

An additional $30 million will go to CanCode, a federal programme funding non-profits that provide digital skills training to young people. The strategy also promises to expand the Global Talent Stream permit programme to accelerate entry for highly skilled AI workers, though specific visa targets are not disclosed.

The question of which jobs AI will create, rather than destroy, remains contested. The strategy’s 250,000-job figure lacks a detailed methodology, and independent economists have not yet validated it.

What comes next

Cross-party pressure is already mounting. A group of parliamentarians from multiple parties has called on the government to block the development of superintelligent AI entirely, arguing that safety guardrails should precede industrial policy. That position finds an unlikely ally in the Vatican, where Pope Leo called for AI to be “disarmed” and rejected just-war theory as “outdated” in the context of autonomous weapons.

Carney has staked his government’s credibility on the claim that sovereignty and safety can advance together. The $2.3 billion bet suggests Ottawa is serious about the first half of that equation. Whether the safety architecture arrives before the supercomputer does will determine whether the strategy earns the moral weight the Pope asked for, or simply the compute capacity investors wanted.