The Trump administration’s sweeping AI Action Plan, released in July, moves the global AI race into a new realm. It’s no longer just a race between OpenAI and Google; it’s a geopolitical contest that the world’s greatest tech power is doubling down on as it seeks to influence the digital decades ahead. The Trump administration boldly declared its ambition for “global dominance” in AI. This is a shot across the bow for other advanced economies with AI ambitions, including Canada.
At its core, the Action Plan signals an unapologetic “America First” tech strategy. It casts AI leadership as a non-negotiable pillar of U.S. economic and military power, tying it explicitly to the values of free enterprise and free expression. The plan is an aggressive blueprint, identifying 90 actions across three strategic pillars: accelerate AI adoption; build AI infrastructure; and lead in AI diplomacy.
America’s AI plan
Among the proposals in the AI Action Plan, five things stand out:
1. Deregulation trumps safety. The terms “deregulation” or “regulation” (the latter often prefaced with “onerous” or “burdensome”) appear 15 times, while the phrase “AI safety” does not appear at all. Washington is signaling that it intends to strip away regulatory friction and wield government purchasing power to seed early deployment. It even threatens to divert federal funds from states with strict AI rules. This sends a strong de-risking signal to the private sector that AI will be an open field, with safety as an afterthought.
2. Jurisdiction. Big data is inherently global and local. And now Trump wants to unshackle Big Tech from state-level oversight of AI. Prime Minister Carney may soon face the same challenges with the provinces as he tries to develop a “one economy” approach to economic development. Both Carney and Trump will face pushback if the big platforms move into health and education data, which are regarded as subnational jurisdictions in both countries. Both leaders will need to figure out how to balance the individual, local, national, and the global in our AI age.
3. Ideology. Trump is aiming to “de-woke” AI models. The issue of ideological bias cuts across party lines, though, so it’s not clear how to do that, especially if we want to avoid some version of a thought police patrolling the algorithms. AI models should be accountable to public standards, including free speech, of course. However, we just don’t know how to temper what we’ve unleashed other than to prosecute developers under the law, just as we do with other forms of speech. Whatever your view, the Trump policy begins a new chapter in the politicization of tech.
4. Investment. A gold rush is underway for data centres, driving billions of investment into AI infrastructure. Trump is laser-focused on keeping and building them in the U.S. Canada can continue to feed that model with our energy, financial capital, and data—or build our own competitive strategy. Data waits for no government.
5. Sovereignty. This may be the most challenging aspect for Canada. The U.S. and Chinese models, and clouds, have become so big and powerful that it’s hard to imagine other countries creating anything to rival them. But there’s a chance for Canada. We have global tech leaders in OpenText, Shopify, and Cohere, and some competitive advantages in our own data sets, especially in health care. Is there a moonshot opportunity to build a Canadian rival? And will that require the same sort of techno-nationalist policies we’re seeing emerge in the U.S. and Europe?
A strategy for Canada
In this moment of great uncertainty with our southern neighbour, Canada is confronted with a strategic question: double down on an AI strategy that is built along national lines, or use AI to strengthen ties with the U.S., integrating our AI strengths into a broader North American innovation bloc? Whatever the case, we see six key steps Canada should take to develop its AI capabilities:
1. Build compute capacity. Canada’s AI Sovereign Compute Strategy must move from announcement to capacity on the ground. Compute is one of the critical components to keeping high-value data, talent, and IP inside our borders. Closing the scale gap with the U.S. means accelerating the construction of domestic data centres.
2. Educate at deployment speed. Executive-level AI literacy, rapid up-skilling programs, and more aggressive digitization incentives are essential if productivity and wage growth are to be competitive. As a country, we will need to pursue initiatives as bold as the White House Executive Order on AI literacy, which mandates K-12 education on AI.
3. Lead with GovTech. Make federal and provincial agencies living laboratories for responsible AI—think big ideas for reduced ER wait times or micro-targeted solutions for faster passport renewal—which can both improve domestic outcomes and provide proven solutions to export.
4. Streamline strategic infrastructure. America declared a “National Energy Emergency”, in part to cut red tape around AI-related energy and permitting. Canada does not need to mirror that approach, but we cannot afford multi-year lags for projects that underpin AI capacity.
5. Anchor Canada as a reliable trade partner. Multinational firms are looking for stable, low-tariff (or tariff-free) production bases inside North America. Canada can pitch itself as the secure R&D and manufacturing node, balancing U.S. market access while also maintaining interoperability with evolving EU privacy norms.
6. Future-proof the workforce. Beyond minting more AI engineers, Canada must help mid-career workers whose roles will change. Scaled-up sectoral programs and portable training supports will decide whether AI augments the broader labour force or leaves vulnerable segments behind.
The U.S. has boldly declared that it wants to dominate the future of AI. Faced with a multi-decade productivity challenge, restricted immigration, workforce transition, and trade frictions, it’s imperative that Canada commits itself to becoming a global AI leader.
This article was made possible by RBC and the generosity of readers like you. Donate today.
