Lawrence Zhang is head of policy at the Information Technology & Innovation Foundation’s Centre for Canadian Innovation and Competitiveness.

Canada’s digital policy is at a crossroads. One path focuses on integrating global, best-in-class digital technologies into Canadian firms and institutions, driving productivity. The other pursues “digital sovereignty:” a costly and likely fruitless attempt to recreate infrastructure Canada already accesses through global markets.

The latter has growing appeal, especially after a year of trade threats and casual insults coming from Washington that have sharpened national sensitivities. But emotion is a poor substitute for strategy.

One example of this framing appears in a recent roadmap from the Information and Communications Technology Council (ICTC), which argues that “enhancing domestic digital capacity is an investment in national sovereignty,” and treats Canadian compute, cloud infrastructure, and semiconductor capacity as critical. 

ICTC is not alone in advancing this view. In recent publications, the Canadian Shield Institute has articulated a vision of digital sovereignty that prioritizes domestic ownership and insulation from foreign control as central tests of competitiveness and security, even when foreign firms operate and invest within Canada.

Start with cloud and AI compute, where the sovereignty argument is most visible. Canada does not lack access to cloud services or advanced computing capacity. Canadian firms can already purchase world-class cloud and AI services through global markets. The binding constraint is not supply. It is demand.

Firms underinvest in digital tools, struggle to integrate new technologies into existing workflows, and face internal capability gaps that slow adoption. Governments themselves remain cautious users of cloud and data-driven systems, reinforcing that hesitation across the economy.

In this context, building “sovereign” cloud or AI capacity does not raise productivity. The problem is not ambition but sequencing – without widespread adoption by Canadian firms and governments, supply-side investments face thin markets and weak signals, making sustained performance and scale difficult to achieve.

The same ownership logic is now being extended to data. Calls for stricter data residency requirements are often framed as necessary for trust, privacy and innovation. In practice, they confuse governance with geography.

What matters for security and research integrity is who can access data, under what rules, and with what enforcement, not whether servers sit inside a particular border. Treating data sovereignty as a location problem raises costs and fragments systems without meaningfully improving outcomes. It slows adoption precisely where speed and scale matter most.

When this logic moves from governance into industrial policy, the costs grow dramatically. ICTC's proposed roadmap makes this shift explicit in its call for greater domestic semiconductor capacity, and semiconductors illustrate exactly why that approach fails.

Leading-edge chip manufacturing is among the most capital-intensive industries on the planet. Competing at the frontier requires tens of billions of dollars per facility, years of accumulated expertise, and enormous scale. Canada has no firm positioned to anchor a viable fabrication ecosystem, and without one, “sovereign” semiconductor manufacturing would require permanent public subsidy. Even then, late entry would almost certainly deliver higher costs and weaker performance.

China, despite vastly greater resources and industrial coordination, continues to struggle at the frontier. A realistic semiconductor strategy for Canada is integration into allied supply chains; not attempts at building fabs the country cannot sustain.

Economic value comes from application, integration and scale

Much of the rhetoric around “becoming a digital leader” rests on a misleading premise: that Canada is sliding from leadership into dependency. Canada was never a leader in general-purpose computing hardware or semiconductors, and it did not need to be.

Economic value has always come from application, integration and scale. The firms and countries that capture value are those that deploy technologies widely and build businesses around their use, not those that insist on owning every layer of the stack.

That reality is already reflected in Canada’s economy. As ICTC itself notes, the digital economy accounts for about 10 percent Canada’s total GDP. Those gains did not come from manufacturing laptops or fabricating chips. They came from integrating global technologies across sectors like finance, logistics, agriculture, energy, and services.

Most advanced economies work this way. Medium-sized countries like Canada become digital leaders by being leading-edge users and turning adoption into productivity.

Across cloud, data, AI and semiconductors, however, the same mistake keeps repeating: treating ownership of infrastructure as a substitute for widespread use. Canada already relies on global systems for payments, navigation, communications and finance. We do not talk about credit-card sovereignty or GPS sovereignty. Drawing a special line around digital infrastructure does not change that reality.

Canada is a trading economy. Importing advanced inputs is not a weakness; it is the global division of labour works. Moreover, you cannot credibly argue for global integration while treating imports as a vulnerability.

The commercialization argument is often invoked to justify greater domestic ownership of digital infrastructure, but it does not hold up. Foreign ownership of digital infrastructure is a convenient but weak explanation for Canada’s poor patent commercialization record.

Innovations commercialize where customers exist, where procurement pulls technologies into use, and where firms can scale quickly. Canadian discoveries benefit foreign economies because foreign markets offer faster access to demand and revenue. Wrapping commercialization in the language of national control does not fix that problem. Ownership rules do not create users. Markets do.

The alternative is neither retreat nor dependency. It is focus. A credible digital industrial strategy would treat adoption as the core objective, not a downstream hope.

When Canadian firms buy, use and integrate advanced digital technologies at scale, they create real domestic customers. That demand is what allows new Canadian technology firms to emerge, scale and survive without permanent subsidy. 

Seen this way, sovereignty is not something you declare or build in isolation. It is something you earn through capability. Economies are secured by productivity, by using the best tools available, within allied systems, faster and more effectively than peers.

Canada’s problem is not dependence. It is underuse.

R$