In a previous post , I traced the arc of Canada's AI governance journey: from the world's first national AI strategy in 2017, through the long and ultimately unsuccessful attempt to pass the Artificial Intelligence and Data Act (AIDA) as part of Bill C-27, to Alberta stepping into the federal vacuum with its own privacy modernization and data centre ambitions. The conclusion, as of late May 2026, was uncomfortable but accurate: Calgary organizations were deploying AI at scale without a single dedicated AI law on the books at any level of government.
The federal government's announcement on June 4, 2026, deserves careful reading. Not just for what it says, but for what it is and what it is not.
What Was Actually Announced
Prime Minister Mark Carney launched "AI for All," framed as Canada's national AI strategy. The headline numbers are ambitious: a projected $200 billion boost to the economy, 250,000 new AI-related jobs by 2031, and a target of raising national AI adoption from roughly 12 per cent to 60 per cent by 2034. The strategy is organized around six pillars: protecting Canadians, empowering skills, driving adoption, building sovereign AI infrastructure, scaling Canadian companies, and building trusted international partnerships.
The initiative includes a National AI Literacy Initiative offering free entry-level AI training for all Canadians, access to trusted AI agents for post-secondary students, a proposed world-leading public AI supercomputer, and expanded support for small and medium-sized businesses to adopt AI tools.
Minister of AI Evan Solomon, who publicly confirmed in June 2025 that AIDA would not be reintroduced in its original form and signalled a preference for something "light, tight, and right," described the new strategy as "putting artificial intelligence to work for Canadians."
Proposal or Law? The Distinction Is Not Semantic
The most important thing for an executive audience to understand is this: "AI for All" is a strategy document and a policy announcement, not legislation.
The government has stated that the strategy will introduce new legislation, investments, and programs. The operative word is "will." No bill has been tabled. No regulatory text has been published. What exists today is a six-pillar framework, a set of stated intentions, and economic targets.
This matters practically. The obligations that matter for compliance officers, legal counsel, and board-level governance committees are those that carry legal force. A strategy document, however well-resourced, does not create enforceable rights, duties, or penalties. Until legislation passes Parliament and receives Royal Assent, organizations face the same patchwork that has governed AI deployment for the past several years: PIPEDA at the federal level, provincial privacy statutes (including Alberta's recently modernized POPA), sector-specific regulation, and the government's existing voluntary Code of Conduct on the Responsible Development and Management of Advanced Generative AI Systems.
The successor to AIDA, which Minister Solomon has signalled is coming and which the "AI for All" strategy presumably contemplates, has not yet been tabled. When it is, it will face the same legislative gauntlet that consumed AIDA for over two years. The path from strategy to binding law in Canada is neither short nor guaranteed.
The Case for the Strategy
Canada's AI adoption rate of approximately 12 per cent is not a point of pride for a country that helped pioneer the field. If the strategy's policy levers (SME adoption programs, skills investment, public sector AI deployment, and sovereign compute infrastructure) actually move that needle toward 60 per cent, the productivity and economic gains would be substantial. The $200 billion GDP projection is aggressive, but not implausible if adoption scales meaningfully across healthcare, financial services, energy, and government.
The sovereignty pillar addresses a concern shared across the political spectrum: that Canada's AI future is increasingly shaped by infrastructure, platforms, and capital controlled outside its borders. A world-leading public AI supercomputer and a commitment to Canadian-governed compute infrastructure, if actually built, would give Canadian organizations real alternatives to foreign dependency.
The consultation process that preceded the strategy carried genuine scale. More than 11,000 Canadians and 28 task force members participated in shaping it. That provides a democratic foundation that critics of the earlier AIDA process argued was missing from that legislative effort.
The Case Against, and the Public Response
The public reaction has been pointed, and executives would do well to understand it rather than dismiss it.
An Angus Reid poll cited in coverage of the announcement found that 68 per cent of Canadians would oppose a large AI data centre built near their home, a number that rises to 73 per cent among rural Canadians. NDP leader Avi Lewis characterized the strategy as a government "imposing a world-changing technology on all of us without any public debate," arguing it prioritizes adoption over protection. Manitoba Premier Wab Kinew rejected a provincial data centre proposal on environmental and economic grounds.
The critique goes beyond anti-technology sentiment. The substantive concern, articulated by critics across the political spectrum, is that the strategy accelerates adoption while the protective measures it promises remain entirely dependent on legislation that does not yet exist. Deepfake protections, surveillance pricing rules, and the proposed online safety regime are commitments on paper, nothing more.
The energy infrastructure question is real as well. AI data centres are significant consumers of power and water. For Alberta, with its stated ambition to attract $100 billion in data centre investment, the tension between economic opportunity and community impact is already visible. Organizations planning infrastructure at scale should expect that tension to grow.
The employment projections also warrant scrutiny. The promise of 250,000 new jobs is set against an economic reality in which AI is simultaneously automating roles across exactly the sectors those jobs would serve. The net employment effect of aggressive AI adoption is genuinely contested among economists. A national strategy that leads with job creation numbers while staying vague on worker transition supports invites legitimate skepticism.
What This Means for Calgary Organizations
The announcement has not changed the immediate compliance picture, but it has sharpened the timeline.
The federal government has now publicly committed to introducing AI legislation. That commitment, combined with Minister Solomon's earlier signals, Alberta's OIPC continuing to push for a standalone provincial AI statute, and Quebec's existing automated decision-making obligations, means the regulatory density I projected in my last post is arriving on a timeline that is now more visible, if still uncertain.
Organizations treating AI governance as a future compliance concern rather than a current operational priority are running out of time. The strategy's protective pillar specifically references modernizing privacy law, regulating deepfakes, and introducing an online safety regime. When that legislation arrives, it will not come with a grace period for organizations that have not yet inventoried their AI systems, documented their risk assessments, or established clear accountability structures.
The sovereignty narrative also has direct relevance to Calgary's energy sector. If federal investment flows into Canadian AI infrastructure and sector-specific applications, and if the government follows through on positioning Canadian companies as preferred suppliers to federal programs, organizations with credible AI governance frameworks will be better positioned to compete for those opportunities. Those without them face a competitive disadvantage on top of a compliance one.
The Bottom Line
"AI for All" is a serious strategy document with credible economic ambition and a political mandate behind it. It is also, at this moment, exactly that: a strategy. The legislation it promises has not been written, tabled, or debated. The investments it announces are subject to budget processes and parliamentary approval. The adoption targets will require sustained execution over years.
Uncritical enthusiasm is not warranted. Neither is reflexive skepticism. Canada's AI governance moment is real. The government's direction is clear. The gap between intent and law remains wide.
Build your governance infrastructure as if the law is coming, because it is. Just don't reorganize your compliance calendar around a Royal Assent date that hasn't been set.