Canada’s “AI for All” Plan Means AI Approved by Ottawa

Canada’s new national AI strategy is called “AI for All.” If we’re being literal, it should be called “AI Approved by Ottawa.”

Released this week by Minister of Artificial Intelligence and Digital Innovation Evan Solomon, his first foray into translating the policy priorities of the Liberal government into action, the strategy commits more than $2.4 billion in new federal spending and folds in another $6 billion from the existing budget.

Industrial Policy with a values filter

Though there is plenty of serious consideration of the energy needs of AI innovation and build-out (where Canada would have a clear strategic advantage), much of the AI plan’s substance is bureaucracy-driven industrial policy with a values filter: a $50 million Creative Technology Program tied to favored creators, an “equity-based national standard” for AI development, and explicit funding mandates for tools that “reflect Canadian identity, culture, and inclusion.”

Rather than embracing competitive forces and setting common-sense rules that have given rise to the growth of AI industries elsewhere, too much of the government’s AI plan rests on handing out government funds and mandating certain tracks of entrepreneurial development. This is far from winning strategy.

For starters, provincial rules on AI are already being drafted and will compound over time. How will the federal government contend with these, not to mention any AI entrepreneur or researcher?

The Equity Filter

The most consequential line in the Canadian document is the call for an “equity-based national standard” governing AI development. A standard like this hands federal officials the authority to decide which AI tools sufficiently advance equity before they receive procurement preference or public investment. It is not a stretch to imagine how this could be abused to favor politically-connected firms with the inside track to funding.

On this metric, Canadian consumers lose twice. Compliance costs rise and get passed through to anyone using federally-supported AI. Products that fail the political screen test are locked out from funds. The result is fewer choices, not better ones, which is the inverse of what the strategy claims to deliver.

Identity-Based Funding

The section labeled “Supporting Indigenous leadership in AI” is honest about what it does: direct funding and leadership roles toward specific identity groups. Framed as inclusion, the mechanism is exclusionary by design. Public money picks which voices count in Canadian AI.

That same logic runs through the new $50 million Creative Technology Program and a $200 million federal health mission, both governed by the strategy’s cultural and equity goals. Government picks the recipients, shapes the output and calls it support for creators. The market test, whether anyone actually pays for the product, is replaced with a political one. Considering the ills of current and ongoing funding for the news and media sector, why would we ever want to see this replicated for AI?

A Canadian shopping for an image-generation model, a coding assistant or a translation tool generally wants the better tool at the lower price. Ottawa wants them to want the more inclusive tool, and is willing to spend public money to nudge them. That isn’t a role we should want for government in the 21st century, and especially not with a technology as revolutionary as AI tools, agents, and the next generation of autonomous computing.

The American Mirror

Compare it to what Washington just did. The Trump administration’s AI Action Plan, released last summer and reinforced by a December 2025 executive order, pushes federal preemption of state AI laws and steers federal procurement toward AI that is “ideologically neutral.”

The United States is trying to dismantle the checkpoints while Ottawa is adding more. A Canadian developer comparing the two regimes can see which side the capital, the talent and the next generation of AI startups will flow toward.

Who Gets the Money?

Now, let’s follow the $2.4 billion. The $500 million Canadian Tech Growth Fund lets Ottawa take equity stakes in AI firms federal officials judge “promising.”

The $500 million Regional Artificial Intelligence Initiative is distributed through Regional Development Agencies, the same political vehicle Liberal governments have long used to steer spending toward favored ridings. The $130 million for commercialization flows to the three established National AI Institutes — Vector, MILA and Amii — incumbents that already shape the federal AI pipeline.

The $100 million early-stage venture fund explicitly targets companies “emerging from Canada’s leading research universities,” locking in the same handful of institutions.

The pattern is consistent. Almost none of this money reaches Canadian developers, consumers or new entrants as broad-based tax relief, regulatory clarity or open infrastructure. It flows through federal agencies, Crown corporations and a small set of pre-approved institutional partners. AI for All is not a market-opening strategy. It is a distribution mechanism for an insider economy that decides who builds Canadian AI before the consumer has any say.

The Cultural Layer

The strategy then layers French-language and broader cultural requirements onto AI development. Supporting French-language tools is reasonable. Canada has roughly 7.4 million Francophones and a real public interest in AI that serves them. But much of the availability of AI tools in different languages will come purely from consumer demand. Tech firms and AI tool builders want to be able to serve consumers across Canada and the world, and they will provide tools in native languages. We do not need to crack open the government coffers to incentivize builders to serve customers in the languages they want.

But embedding cultural mandates into technical standards and procurement is a different matter. It risks fragmenting the Canadian AI ecosystem precisely as global competition heats up. Quebec already enforces its own AI-relevant privacy regime under Law 25. Layering federal cultural rules on top compounds the load on every Canadian developer, exactly when the US is using federal authority to clear the regulatory field instead.

What’s Missing

For all the spending, the gaps in the document are as telling as its mandates. There is no commitment to open-source neutrality, no regulatory sandbox for early-stage AI products, no consumer data rights framework specific to AI systems and no signal of how Ottawa will keep provincial AI rules from multiplying the compliance maze. The consumer never appears as a stakeholder with rights, only as an audience for approved content.

A consumer-first strategy would look different. It would set open standards for interoperability. It would create a federal sandbox letting startups test new products under provisional rules. And it would draw a clear preemption line to head off a patchwork of provincial AI mandates before it forms.

On the energy front, it is a big missed opportunity. As a cold-climate country with affordable electricity, Canada has a natural advantage to be the location for massive amounts of AI compute.

Will data center investors and builders get support from Ottawa to streamline and quickly build projects without mountains of red tape and bureaucratic overhang?

Will only favored projects get the green light to move forward if they accept Ottawa’s demands on their energy usage and which percentage will be sustainable? Will green NGOs get first veto rights over projects, or will the government make it easier for builders to build?

This isn’t addressed too finely in the report, and it could have been a big sell for international investors and developers.

Ottawa chose another path. “AI for All” sounds inclusive. In practice it means AI that passes Ottawa’s test, built by approved hands, sold to a consumer who never asked for any of it.

Published on Consumer Choice Center.