Carney government releases AI road map that aims to make Canada a leader

“The question is, ‘Will it improve the lives of all Canadians or benefit only a few?’ And that’s why we must take a positive, pragmatic, and prudent approach that builds safe, reliable, and sovereign AI for workers and businesses, for Canada, and for our allies.”

The prime minister reiterated one of the strategy’s core messages: that while Canada is viewed as an AI pioneer with a trustworthy reputation, the country suffers from a serious adoption gap, with few businesses making use of the technology and low global rankings when it comes to AI training, literacy, and trust.

“We’re a pluralistic society that works. We’re a stable, reliable partner in a world that is anything but, but first we have to be honest about the risks that AI poses to Canadians, and the challenges that Canada faces,” said Carney, citing the dangers posed by deepfakes, unsafe chatbots, disinformation, and privacy threats.

“We are highly dependent on foreign suppliers for the infrastructure that powers AI, from compute to cloud to data storage. That creates real risks that foreign entities could access Canadian data, deploy AI products that shape Canadian lives, without reflecting our values, and tilt the playing field against Canadian firms, while Canada lacks the leverage to push back.”

Thursday’s strategy, which commits around $2.3 billion in new funding, is intended to account for advances in AI since former prime minister Justin Trudeau’s days in power. It also supports Carney’s ambitious artificial intelligence agenda, which is focused on adoption rather than the regulatory constraints favoured by his predecessor.

The sprawling document is based on six pillars, targets five priority sectors, highlights 11 future outcomes, and outlines 57 key actions.

Its top pledges include:

• Creating up to 90,000 “AI-related” jobs, including 45,000 positions for young Canadians, along with another 250,000 jobs through the adoption of AI, by 2031.
• Boosting business adoption of AI from 12 per cent today to 60 per cent by 2034.
• Launching a new “AI missions” program to solve national challenges in key sectors like energy and agriculture, beginning with a $200 million health care project to reduce ER wait times, expand access to primary care, and lower administrative burdens.
• Introducing a $500 million Canadian Tech Growth Fund to help Canadian AI companies scale up their operations and remain within the country, with those companies also benefitting from money stemming from the previously-announced Sovereign Wealth Fund.
• Expanding the existing Compute Access Fund to $700 million to give small- and medium-sized enterprises access to affordable computational power and resources.

But the strategy’s objectives, Artificial Intelligence Minister Evan Solomon told the Star, are subject to change.

“If there’s new opportunities or new challenges, security challenges, we are going to adjust, and we are going to make sure that we have the flexibility to do so. This a very comprehensive start, but we are not pretending … that you can handle AI in one document and chip it in stone,” Solomon said.

Nowhere are those challenges more acute than striking the balance between relying on domestic or foreign computing.

The strategy notes that “Canada will require 5.5 gigawatts worth of AI compute for its commercial players by 2030” and that much of that power will come foreign “hyperscalers” – companies with vast computing infrastructure and data storage capabilities.

“This is not an adversarial conversation…we’re very much working together as a very complementary reinforcing,” Carney said, when asked about the risks of depending on the U.S. for that power.

The prime minister said that nevertheless, there will be “certain elements where for national security for Canadians’ preferences and for sovereignty” the goal will be to prioritize infrastructure here at home. 

The document positions Canada as a leader of a coalition of middle-power democracies that could “offer a credible alternative” to dominant market actors.

“We have LLM — large language model — capabilities here, that very few other countries, only three other countries have, that’s part of a broader AI ecosystem that we can build alliances around,” Carney said, citing the government’s existing Sovereign Technology Alliance that has already resulted in a partnership with Germany.

The rest of the strategy is largely aimed at improving the adoption, trust and understanding of the technology. 

Solomon told the Star that February’s Tumbler Ridge tragedy, in which it emerged that the shooter had previously used OpenAI’s ChatGPT in a concerning manner, injected urgency into the safety aspect of the framework.

But aside from referencing expected legislation on online harms and modernizing consumer privacy law, along with reviewing the Privacy Act through a digital lens, there is little in the blueprint that clearly explains how AI companies and their tools will actually be held to account, despite Carney’s insistence Thursday that regulations and protections, particularly for child safety and privacy concerns, are on the way.

The strategy notes that Ottawa will look at requiring that artificially-generated content be branded with watermarks, and that Canada will ensure that personal information is not used for purposes such as surveillance pricing.

The road map also proposes a national AI literacy initiative and pledges that one million entry-level post-secondary students will have access to free materials, though few details for those projects were revealed. 

The government is also striving to give all undergraduate students access to “trusted AI agents” to equip them for the workforce.

Government officials, speaking on a background basis to reporters prior to the strategy’s release, were unable to explain what tools would be provided to students and what the cost of that undertaking would be, only saying that details would be shared in the weeks ahead.