Carney Launches ‘AI for All’ Targeting $200 Billion in AI Growth

Canada’s business AI adoption sat at just over 12 percent in 2025, one of the lowest rates in the G7, when Prime Minister Mark Carney launched “AI for All” in Toronto on Thursday and pledged to push that figure to 60 percent by 2034 and unlock $200 billion in additional economic growth. The plan, the most comprehensive AI commitment in Canadian federal history, targets 250,000 new AI-related jobs over five years and at least $2 billion in new investment across infrastructure and workforce training.

The strategy runs to 2034 and was shaped by national consultations in 2025 that drew more than 11,000 submissions. A global survey conducted the same year placed Canada 44th out of 47 countries in AI literacy and 42nd in trust in AI systems, rankings the strategy explicitly calls “binding constraints” on adoption.

Canada’s AI Adoption Gap at 12 Percent

Statistics Canada puts the precise figure at 12.2 percent of firms that integrated AI into their production or service delivery in 2025, doubling from the prior year but leaving Canada’s business AI adoption share below the OECD average, at 20th out of 35 member countries. The strategy document cites Nordic countries, Germany and France as the benchmarks Canada needs to reach.

The SME Digital Maturity Gap

The gap concentrates in small and medium-sized enterprises (SMEs). The Business Development Bank of Canada (BDC) assessed digital maturity across five dimensions and found only 8 percent of Canadian SMEs at a very high level, with fewer than a quarter reaching high or above. BDC estimates that moving all Canadian SMEs to very high digital maturity could lift SME productivity by 38 percent, translating under favorable conditions to roughly $350 billion in GDP.

• 12.2% – share of Canadian firms using AI in production or services in 2025 (Statistics Canada)
• 20th – Canada’s rank among 35 OECD nations in AI adoption (BDC)
• 8% – Canadian SMEs at very high digital maturity (BDC)
• 38% – estimated SME productivity lift if all reached very high maturity (BDC)

Regional divides compound the picture. Sage’s survey of 2,012 Canadian SME leaders found adoption running from 56 percent in Quebec to just 28 percent in Northern Canada, a disparity the report calls a “two-speed digital economy.” Finance and technology firms are considerably further ahead than agriculture and construction.

The Imagination Gap and Its Evidence

The RBC report on Canada’s AI imagination gap identifies the core challenge as a widespread inability among Canadian business leaders, especially at smaller firms, to see where AI is relevant to their own operations. RBC found Canadian firms also explore a narrower set of AI use cases than firms in comparable OECD economies, a pattern OECD data tracks consistently. The Canadian Federation of Independent Business (CFIB) found that SMEs already using generative AI tools gain an average of 2.05 hours per day for every 0.97 hours they invest. Statistics Canada found an additional 14.5 percent of Canadian firms planned to adopt AI within the following 12 months, suggesting the technology is clearly on business owners’ radar even where action hasn’t followed.

What Carney Announced at Toronto General

Carney chose Toronto General Hospital and the University Health Network (UHN) for the Thursday launch, a venue where AI is already working in diagnostics and patient care. Evan Solomon, Minister of Artificial Intelligence and Digital Innovation, stood alongside him. The strategy was informed by 11,000 submissions and a 28-member AI Strategy Task Force spanning industry, academia, unions and community groups.

The Core Targets

The full commitment list appears in Ottawa’s official AI for All strategy announcement. The $500 million Canadian Tech Growth Fund would allow Ottawa to take equity stakes in domestic AI startups, using government procurement as an anchor customer for early commercial-stage companies. A National AI Literacy Initiative delivers free entry-level training through libraries and community organizations. Every post-secondary student in the country gets access to a “trusted AI agent,” in Carney’s phrasing.

Regulation, Talent, and the First Mission

The first AI Missions Program focuses on health care, backed by $100 million to launch the Health Sector Data Space, which links secure and standardized health datasets for clinical trials and AI diagnostics. The strategy also expands the Canadian Institute for Advanced Research (CIFAR) AI Chairs program and opens an accelerated entry pathway for highly skilled international AI workers through the Global Talent Stream.

On regulation, the strategy commits to updating personal information legislation with specific focus on “surveillance pricing” (where algorithms use personal data to set individualized prices for identical products) and creating a new online safety framework covering AI chatbot and social media platform users.

Valérie Pisano, chief executive of Mila, the Montreal AI institute, called the strategy “ambitious” and said it “puts a stake in the ground” for Canada’s AI ambitions. Conservative deputy leader Melissa Lantsman panned it, saying her party doubted the Liberals could deliver AI jobs for young people given broader economic pressures.

The $890 Million Supercomputer Bet

The infrastructure component of Thursday’s announcement builds on a program already running. In April 2026, the government launched the AI Sovereign Compute Infrastructure Program (SCIP), a competitive call for applications to design, build and operate a national public AI supercomputer. The SCIP program page shows applications closed June 1, three days before Thursday’s launch. The program allocates $890 million to its Infrastructure Build Layer over seven fiscal years starting in 2026-27.

The strategy’s description of the current infrastructure situation is pointed. Canada’s AI data centres and cloud offerings are “largely foreign-owned and controlled,” with sovereign compute capacity described as “nascent, particularly in cloud.” Any contractor selected through SCIP must deliver a functioning system within 18 months of signing a contribution agreement. The program is structured in two integrated layers: an Infrastructure Build Layer that designs and operates the hardware, and a National Service Layer handling user support, training and data service integration.

Alongside the supercomputer, Canada will partner with private capital to build data centres scaling to at least 100 megawatts, aligned to clean energy goals. Carney put the sovereignty rationale plainly at Toronto General: “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.”

The SCIP extends a Canadian Sovereign AI Compute Strategy with three pillars: mobilizing private sector investment, building public supercomputing infrastructure, and running an AI Compute Access Fund. A 2024 investment of $2 billion in compute capacity supported upgrades at all three national AI institutes. Canada has run a strong AI research program through the Vector Institute in Toronto, Mila in Montreal, and the Alberta Machine Intelligence Institute in Edmonton. The sovereign infrastructure push closes the gap between that research record and commercially deployed capacity at national scale.

A Country Ranked 44th for AI Trust

The strategy was drafted with specific evidence in front of it. A 2025 KPMG International and University of Melbourne study on global AI trust surveyed 48,000 people across 47 countries, including 1,025 Canadians, and found Canada ranked 44th in AI training and literacy and 42nd in trust in AI systems.

The Canadian-specific numbers break down precisely: 24 percent of respondents had received any AI training, against 39 percent globally; 38 percent reported moderate or high AI knowledge, compared with 52 percent worldwide; 34 percent trusted AI-generated information, versus 46 percent globally. Half of Canadians approved or accepted the use of AI, against 72 percent worldwide. The strategy calls these gaps “binding constraints” on the technology’s growth in Canada.

A 2026 KPMG analysis found 46 percent of Canadians believing AI’s risks outweigh its benefits, and 56 percent of employees said concerns about AI hallucinations (where models produce false or misleading outputs) limited their use of AI at work. The same survey found Canadians clear about what governance conditions would change their minds: 89 percent said laws should control AI-generated misinformation, and 83 percent said they would trust AI more if human mechanisms to review and override its outputs existed. Only 51 percent of Canadians expressed confidence in commercial organizations’ ability to develop AI responsibly, while 79 percent trusted universities and research institutes to do so.

Canada’s personal information legislation, written before large-scale AI deployment was possible, is due for comprehensive reform under the strategy. Ottawa introduced the Protecting Victims Act in December 2025, prohibiting non-consensual sexual deepfakes and increasing penalties for image-based abuse. Broader statutory reform covering how AI companies collect and use Canadians’ personal data has no published legislative timetable in the current document.

Productivity Returns and the Strategy’s Evidence Gap

Vass Bednar, managing director of the Canadian Shield Institute for Public Policy, reviewed the document and was direct about what she found missing.

This strategy takes it on faith that AI adoption will lead to productivity gains and economic prosperity, but it’s not totally clear what the economic value capture strategy is.

Bednar added that the document “seems to be mostly interested in adoption, and then worrying about the tough stuff later,” and flagged a gap in intellectual property protections for Canadian AI developers.

The critique has empirical grounding. Research firm Sage, drawing on its survey of 2,012 Canadian SME leaders, found AI adoption accelerating but “progress towards increased productivity stalling due to barriers around skills, support and costs.” Organizations that converted AI tools into real output gains had built clean data pipelines, restructured workflows and established written AI governance policies before they scaled. As of early 2026, only 29 percent of Canadian employees could confirm their employer had a comprehensive AI use policy, with 42 percent unsure whether one existed, per KPMG’s March 2026 analysis.

A Statistics Canada study published in April 2026 found AI adoption growing but productivity gains dependent on what it calls “complementary capabilities”: data quality, workforce skills and organizational process redesign. Firms with stronger foundational capabilities recorded larger productivity returns from the same AI tools.

The BDC data runs in a different direction: 97 percent of Canadian SMEs that have already adopted AI report tangible benefits. Ottawa is betting that early adopters are the proof subsequent waves need to see. The strategy targets 250,000 AI-related jobs by 2031 and the 60 percent adoption mark by 2034.

Ottawa’s more immediate test is concrete: select a SCIP contractor this year, and move the $500 million Canadian Tech Growth Fund from announcement to deployed capital. The personal information legislation the strategy promises has no published timetable.