What Canada’s New National AI Strategy Means for Your Business

Canada’s national AI strategy, AI for All, was launched on June 4, 2026. It sets a target of 60% business AI adoption by 2034 and backs it with a $500 million BDC financing program, $700 million in sovereign compute for smaller firms and the $300 million AI Compute Access Fund. It does not include a federal AI law, so PIPEDA and provincial rules still decide what you may do with AI.

Updated October 30, 2026. Corrected the AI Compute Access Fund status: ISED lists it as closed to applications. Added what has launched under the strategy since June (the regional AI initiative, the Amii expansion and the Cohere investment). Earlier update, October 28, 2026: added the launch details, the adoption target and the current Statistics Canada figures.

The Canada national AI strategy deserves the same reading you give a budget. Some of it is real money for your business. Some of it is a brochure. Telling the two apart will save you a budget cycle. Plan around the funded half and ignore the rest until it has a price. Prime Minister Mark Carney confirmed on June 1, 2026 that the long-delayed strategy would be released within days, and the government released it on June 4, 2026 as AI for All. Drafts shared with media and a string of ministerial announcements through May had already made most of the major moves public. Below is what the strategy does for Canadian business owners, what it asks of you, and where it leaves a gap.

Key points

• The strategy is built on six pillars, including powering AI adoption and scaling Canadian AI businesses.
• A new Canadian Tech Growth Fund will let Ottawa take equity stakes in domestic AI companies.
• $66 million has already gone to 44 Canadian companies through the AI Compute Access Fund as of May 12, 2026.
• Free AI literacy training is on the way, with a target of reaching one million post-secondary students by 2031.
• Federal AI regulation is still missing. AIDA died in January 2025. PIPEDA is still your operating rulebook.

What is in Canada’s national AI strategy?

The federal AI strategy is organized around six pillars. The Department of Finance previewed them in the spring economic update, and the draft documents seen by CBC and BetaKit pointed to them, and the strategy released on June 4 kept all six. The pillars are:

1. Protecting Canadians and safeguarding democracy
2. Empowering Canadians with AI training and education
3. Powering AI adoption for shared prosperity
4. Building a sovereign Canadian AI foundation
5. Scaling Canadian AI businesses
6. Building trusted partnerships and global alliances

Three of those pillars carry real programs you can already touch. The other three set direction. Expect a year or longer before they produce anything you can put on an invoice, which is an estimate, and it would be wrong if funded programs appear under those pillars sooner.

What does AI for All promise Canadian businesses?

A fivefold jump in adoption, paid for with loans and cheaper compute rather than grants. Ottawa wants business AI use to climb from roughly 12% in 2025 to 60% by 2034. Statistics Canada put the figure at 19.2% in the second quarter of 2026, so the curve is moving, but the target still needs about three times today’s rate.

The money that reaches an operator comes through three doors. The BDC LIFT program offers $500 million in financing for AI adoption, which means a loan you repay, not a cheque you keep. A further $500 million expands the Regional Artificial Intelligence Initiative. And $700 million in sovereign compute is aimed at smaller firms that cannot buy capacity on their own. The rule for a business owner is simple: ask BDC what LIFT eligibility looks like before you sign a vendor contract, because the financing terms may change which tool is cheapest.

Which programs can a business use today?

The AI Compute Access Fund has paid out, but it is not taking applications: ISED lists it as closed, and its last call ended on July 31, 2025. For a Canadian firm that needs compute now, Canada’s new EuroHPC membership and its JUPITER access rules are the next place to look, though the startup fast lane is not yet open to Canadian applicants. On May 12, 2026, Minister Evan Solomon announced $66 million for 44 companies from the $300 million fund, which pays 50 cents on the dollar for non-Canadian compute and 67 cents for Canadian compute. The details of each program follow.

The AI Compute Access Fund has funded 44 companies but is closed to new applications

On May 12, 2026, Minister of Artificial Intelligence and Digital Innovation Evan Solomon announced $66 million in support for 44 Canadian companies through the AI Compute Access Fund. The total program is $300 million. The fund offsets the cost of high-performance computing, paying 50 cents on the dollar for non-Canadian compute and 67 cents on the dollar for Canadian compute. Companies funded so far work in life sciences, health care, energy, advanced manufacturing, agriculture, finance, natural resources and transportation. Examples include SenseNet, which uses AI to detect wildfires earlier, and Spare, which optimizes public transit routing. Additional funding offers will keep coming as more applications are assessed. If your business is building AI products or scaling internal AI workloads, this is the most concrete program available to you today.

A Canadian Tech Growth Fund will take equity in AI firms

The strategy proposes a Canadian Tech Growth Fund that will hold direct government equity stakes in Canadian AI companies. It targets smaller deals than the Canada Strong Fund, the $25 billion sovereign wealth fund announced April 27, 2026. For founders, this changes the cap table conversation. For operators, it points to where federal procurement and partnership attention will go next.

Sovereign AI infrastructure is being built in Canada

The government and TELUS are advancing a proposal for a large-scale AI data centre in Kamloops, B.C., plus two new facilities in downtown Vancouver and Mount Pleasant. This is the supply side of “use Canadian compute and get a bigger subsidy.” Sovereign compute matters for data residency, contract compliance, and the cost line on your AI roadmap.

Free AI literacy training is coming

The strategy commits to free AI literacy training for Canadians, with a target of reaching one million entry-level post-secondary students and creating up to 90,000 AI-related jobs and work opportunities for young Canadians by 2031. If you hire entry-level talent in Canada, this is the cheapest training pipeline on offer this cycle. Build a co-op or internship pipeline now and pull from it through 2027.

What AI for All has delivered since June

Five months after launch, the strategy has moved from pillars to programs, and the pattern is consistent: Ottawa funds the supply side and the intermediaries, while the business still has to apply, qualify and repay. Three October instruments show the shape of it. On October 22, 2026, PrairiesCan opened its share of the expanded Regional Artificial Intelligence Initiative, which now carries $700 million in federal funding nationally, with Prairie applications accepted until December 31, 2030 or until the money is committed. We worked through what that RAII money is worth to a business separately. On October 27, Amii said a further $2.5 million from PrairiesCan will fund projects with 25 new small and mid-sized Prairie businesses, bringing federal support for that program to $5.6 million. Our view on which Prairie firms should apply for an Amii seat is a separate column. On October 29, the Prime Minister announced a federal commitment to an equity investment of up to $1.6 billion in Cohere, the Toronto model company, alongside private investors. That is the sovereign foundation pillar acquiring a balance sheet, and what the Cohere stake changes for Canadian AI buyers is narrower than the headline suggests. The funding instruments are now real. The regulatory pillar is not.

What is still missing from the strategy?

The strategy does not solve the federal AI regulation problem. Bill C-27, which carried the proposed Artificial Intelligence and Data Act, died on the Order Paper when Parliament was prorogued in January 2025. The new strategy promises a regulatory framework, but no replacement bill has been tabled. That leaves your business running on PIPEDA, a law written in 2000, plus Quebec’s Law 25, Ontario’s Bill 194 and any cross-border rules that come with your contracts. If you sell into Europe, the EU AI Act is still the binding floor for risk-classified systems. The strategy also does not put hard dates on adoption metrics or productivity outcomes. It sets a 2031 horizon for jobs and literacy, but no quarterly milestone you can plan against.

What should a Canadian business do next?

Start with your compute bill. If you train models or run inference, price the next 12 months of spend on Canadian and non-Canadian clouds, so you are ready if the AI Compute Access Fund opens a new call. Its 67-cent Canadian rate would be the better offer. Then weigh federal equity, update your AI governance policy, build a training pipeline and plan where your data will live.

1. Get a clear picture of where your AI compute spend goes. If you are training models or running inference workloads, price your next 12-month compute bill now. The AI Compute Access Fund is closed, but if a new call opens, its 67 cent Canadian compute rate is the better economic offer.
2. Decide if the Canadian Tech Growth Fund changes your capital plan. If you are an AI company raising in the next six months, federal equity is now an option to weigh against private capital.
3. Update your AI governance policy this quarter. With no federal AI act, your contractual obligations, Quebec Law 25, and Ontario Bill 194 are the real bar. A two-page internal policy is the minimum.
4. Build a literacy pipeline. Talk to a local college or polytechnic about the federal AI training programs coming online and reserve early co-op spots.
5. Plan your data residency. The sovereign AI infrastructure push will shape vendor pricing through 2027. Lock in Canadian-region commitments where data sensitivity demands it.

The skeptic’s view

There is a reasonable case for not changing anything based on this strategy. Strategies are not budgets. Equity funds take time to operationalize. The Canadian Tech Growth Fund has no published rulebook, no announced GP, and no public investment criteria yet. The Canada Strong Fund is similarly early. Past federal AI plans, including the Pan-Canadian AI Strategy launched in 2017, are still mid-rollout. Real-world impact for most small and mid-sized businesses has been thin. Some economists argue that paying public money to subsidize AI compute does not change the productivity story unless adoption depth changes first, which is what KPMG‘s November 2025 survey of 753 Canadian business leaders has been flagging for a year. A sober CFO read says: apply to the Compute Access Fund if a new call opens and it fits, ignore the rest until program rules drop.

What to watch

• A published rulebook for the Canadian Tech Growth Fund, with its fund operator, investment criteria and cheque size.
• LIFT eligibility rules from BDC and the first approvals under the expanded Regional Artificial Intelligence Initiative.
• A new AI Compute Access Fund call from the $300 million program. The last call closed on July 31, 2025.
• Each quarterly Statistics Canada reading on business AI use, the only honest scoreboard for the 60% target.
• The promised privacy and online safety bills. Until one is tabled, PIPEDA, Quebec’s Law 25 and Ontario’s Bill 194 set your floor.

Closing analysis

The parts of the strategy a business can use have mostly shipped as loans and compute subsidies. The compute fund is closed to new applications, LIFT is financing rather than free money, and the law that would tell you what responsible AI use means is still missing. Treat the rollout as a procurement window. Map each program against a real workload, and let the press releases go by.

Frequently asked questions

What is Canada’s national AI strategy?

It is AI for All, the federal plan launched on June 4, 2026. It rests on six pillars, from safety and training to sovereign compute and scaling Canadian AI firms, and targets 60% business AI adoption by 2034.

Does Canada have an AI law for businesses?

No federal AI act is in force. The Artificial Intelligence and Data Act died in January 2025, so PIPEDA, Quebec’s Law 25 and Ontario’s Bill 194 are the rules that apply to most business AI use.

What AI funding can a Canadian small business get?

The main routes are BDC’s $500 million LIFT financing, the Regional Artificial Intelligence Initiative and the AI Compute Access Fund, which covers part of high-performance computing costs when a call is open. Its last call closed on July 31, 2025. LIFT is repayable financing.

What is a sensible Canadian SMB AI strategy under the plan?

Pick one costly workflow, run a four-week pilot with one number to measure, and check LIFT or regional funding before you buy. Write a one-page data rule for staff first.

Related reading

• Why Only 2% of Canadian Businesses Are Making Money From AI
• BDC Just Put $500 Million on the Table for Canadian SME AI Adoption

Related reading: four federal AI training routes for Canadian staff.

Disclosure

The author has no relevant financial, advisory, or board relationships with any party named in this column.