“When Trump speaks, we must distinguish between the signal and the noise,” he says. And the signal? Trump wants “an economic union that is the most powerful one on Earth.”

O’Leary is all in, insisting that integration could strengthen Canada without compromising sovereignty. “No one’s going to trade sovereignty, no one wants to be sold,” he says. But is that true, can we get integration without losing sovereignty—especially in the Age of AI?

To answer that, we start with the AI economy’s foundation: data centres.

Data Centres: The Backbone of AI

AI data centers are massive warehouses filled with microchips. They provide the processing power needed to train and run AI. These facilities are now among the biggest and most expensive infrastructure projects on the planet.

In effect, data centres are to the AI economy what roads, bridges, and factories were to the Industrial economy. Governments around the world are racing to attract these facilities, and Canada is no exception.

Alberta aims to become the location of choice for North America. O’Leary and Smith are already working to achieve this by building the world’s biggest AI data center in Grande Prairie.

It’s an ambitious project that highlights Alberta’s strengths; it also raises deep questions about how integration in the AI economy could impact sovereignty. First, a bit of historical context.

Hewers of Wood and Drawers of Water

Canada’s natural wealth—from pelts and fish to lumber and oil—has always been its strength. Yet for much of the country’s history, foreign companies were the ones profiting, shipping resources abroad to create products.

When factories were built here, it was mainly to assemble goods to sell to Canadian consumers, such as cars, appliances, and shoes.

This economic model had benefits and costs. If it brought good jobs and industrialization, it also made Canada dependent on foreign innovation and unable to diversify its economy. Basically, Canadians built a “branch plant” economy, where decisions about what to build and how to innovate were made elsewhere—first in Britain, then in the US. Whatever we call this dependency—a loss of sovereignty or just the cost of doing business—it defined Canada’s development.

Free Trade was seen as a solution to this problem. Giving Canadian businesses access to American markets—deeper integration—was supposed to empower them to build and sell innovative products. And by most measures, Canada has done well—apparently, too well. Trump now sees Canada’s trade surplus as evidence that the agreement is unfair, and so he is threatening to impose tariffs to restore the balance.

O’Leary may claim that Trump wants integration, but we already have it—and he seems to be rejecting it. Trump’s real message isn’t about integration; it’s that he knows how dependent Canada is on this relationship—and that he intends to use this to put America First.

The lesson is clear: economic integration creates opportunities—O’Leary is right about this—but it also deepens dependency and, given America’s size, that leaves Canadians vulnerable to America’s political whims. So, how might deeper integration work in the AI economy?

A Golden Opportunity?

Microsoft President Brad Smith believes the AI revolution is creating new international AI markets and that there is a "Golden Opportunity" for America to capture them. America, he states, has the technology to build the products to serve these markets, but it needs a "smart international strategy" to ensure the products are exported to allies and friends.

Microsoft has a strategy. It plans to invest $40 billion in foreign data centers in this fiscal year alone. The goal is clear: support AI exports. In effect, Smith’s strategy aims to create “AI branch plants” in global markets—including Canada—then use them to export Microsoft’s products.

This is not just a rerun of the movie Canadians have already seen. Note that these products are nothing like automobiles or stereos; they are not just machines, but intelligent machines that will interact with—and even befriend—their users, both personally and collectively.

Take education. Within a decade, every student will have an individual AI tutor, designed to assess and adapt to an individual student’s needs. The AI will learn about their aptitudes, skills, and motivations to engage them personally and help shape their development.

This highly personalized approach will be used everywhere, from healthcare and finances to legal advice and entertainment. AI will be intimately engaged in our personal lives and families. Collectively, it will help us plan and manage our businesses, communities, and governments.

The values, perspectives, strategies, and goals that AIs bring to these tasks will be decided by their designers, not their users—and whatever O’Leary says, that raises critical questions about our independence.

Back to O’Leary: A Test of Sovereignty

In sum, if Canada is now seeking massive foreign investment to build AI data centers, it should come at this with eyes wide open by publicly vetting two looming questions: How much control do we need to ensure our independence, and how much do we care, as Canadians?

As for O’Leary, his plans for the Grande Prairie data center—pitched as the world’s largest—will be a test of whether Canada can balance economic integration with sovereignty in the Age of AI. Will Daniell Smith insist on appropriate provincial controls, or will this project deepen Canada’s legacy of dependency in the Age of AI?

Don Lenihan PhD is an expert in public engagement with a long-standing focus on how digital technologies are transforming societies, governments, and governance. This column appears weekly. To see earlier instalments in the series, click here.