AI Sovereignty in Canada: Who Will Benefit From the Economic Value of Artificial Intelligence?

Date:

Share post:

Vancouver-based systems strategist Sonali Sharma has spent two decades working across social enterprise, public-sector innovation and technology commercialization, including directing strategy for British Columbia’s technology commercialization programs. In her new book, The 90% Code: Rewriting Power for the AI Era https://www.the90code.com/ she argues that Canada’s AI debate is overlooking a fundamental question: Who will own the value this technology creates?

Canada is investing in AI adoption, infrastructure and homegrown companies. But Sharma warns that Canada could build more domestic capacity while its businesses become dependent on foreign AI vendors, its workers train systems that weaken their own bargaining power and its productivity gains remain concentrated among a small number of owners.

For Sharma, AI sovereignty must mean more than keeping data and intellectual property in Canada. It must also ensure that Canadian organizations retain their knowledge and decision-making capacity—and that workers share in the prosperity AI creates.

In an interview with Retail Insider, Sharma talks about the emerging technology in Canada.

What does AI sovereignty mean in economic terms, and where is Canada falling short?

AI creates value at four levels. First, the infrastructure: chips, computing power and data centres. Second, the AI models themselves. Third, the knowledge that makes those models useful: the data, expertise and processes of the organizations using them. And fourth, the economic gains from putting AI to work.

Canada has made real progress on the first two. Our sovereign compute program, for example, requires Canadian governance and control of publicly supported infrastructure. That matters.

But most Canadian businesses, hospitals and public institutions will use AI rather than build it. For them, economic sovereignty also means controlling their knowledge, having a say in how AI is used and retaining a meaningful share of the value it creates.

That’s where I see a gap. We measure AI investment and adoption far more readily than we measure who captures the economic gains.

A Canadian company could adopt AI successfully, increase productivity and still become more dependent on foreign technology providers. Its costs may fall, but some of the value could flow to vendors and investors elsewhere.

Canadian control of technology and shared Canadian prosperity are not the same thing. We need to pay attention to both.

What are the risks of dependence on foreign AI vendors, and what should companies do?

The biggest risk is that a business invests heavily in AI but becomes less capable of operating independently.

Consider a retailer. Its competitive advantage comes from years of understanding its customers, suppliers, pricing and local markets. Much of that knowledge lives with its people.

When a retailer introduces AI into those operations, it needs to know what happens to that knowledge. Is the data protected? Who controls the systems built around it? Can the retailer move to another provider without losing the capabilities it has developed?

Enterprise AI contracts often provide strong data protections. But ownership of data alone doesn’t eliminate dependence. Once critical operations are tied to a particular vendor, switching becomes expensive and difficult.

There’s another risk that gets less attention. If employees increasingly rely on AI to make judgments they previously developed themselves, the business may gradually lose some of its internal expertise. That makes it harder to challenge the technology or operate without it.

My advice is to treat AI as a strategic capability decision, not just a technology purchase.

Protect your data and intellectual property. Negotiate portability and exit rights before committing. Invest in your people’s ability to understand and challenge AI outputs.

The goal should be to make the business more capable, not simply more automated.

How could AI adoption affect workers, and what needs to change?

Workers are not just affected by AI. Their knowledge helps make it valuable.

AI builds on decades of human knowledge, and organizations are now using the expertise of their own employees to make these systems more useful. Yet the people whose knowledge contributes to that value may have little claim on the resulting productivity gains.

For generations, expertise gave workers bargaining power. You developed skills, gained experience and earned a living from what you knew.

AI is beginning to change that relationship.

Recent research by Google DeepMind examines what could happen as AI makes certain forms of expertise reproducible at very low marginal cost. It raises important questions about wages, employment and how organizations will be structured. The outcomes aren’t inevitable, but the risk is real: if expertise becomes easier to replicate, workers may lose some of the bargaining power that expertise once gave them.

This is why I don’t believe retraining alone is enough.

We absolutely need to help people develop new skills. But a retrained worker is still dependent on wages, even as a growing share of economic value may flow to those who own the technology and the businesses deploying it.

We need more ways for workers to share in the gains. That includes profit-sharing linked to productivity improvements, employee ownership opportunities and meaningful involvement in decisions about how AI changes their work.

This is central to my argument in The 90% Code: Rewriting Power for the AI Era. If we are redesigning how work gets done, we should also rethink who benefits from what it produces.

How should governments approach public AI investments differently?

Canada has been a global leader in AI research, supported by decades of public investment. That research helped create enormous commercial opportunities, many of which have been developed and scaled by companies outside Canada.

The lesson isn’t to invest less. It’s to think more carefully about what Canadians receive in return.

I see three opportunities.

First, where public funding supports significant commercial ventures, governments should consider ways for taxpayers to participate in the upside. That could mean equity, royalties or other arrangements suited to the investment. Ottawa’s recent announcement of an equity investment in Cohere is an important example.

Second, public programs supporting AI adoption should consider how the productivity gains are shared. Businesses that invest in their workers, introduce profit-sharing or create employee ownership opportunities could receive stronger incentives.

Third, we need better measures of success. Investment attracted and adoption rates matter, but they don’t tell us enough about the economic benefits. Are wages improving? Are Canadian companies retaining valuable intellectual property? Are more people participating in business ownership?

Public investment should create public value, and we should be able to demonstrate where that value goes.

Matheus Bertelli photo
Matheus Bertelli photo

What practical steps should Canada take over the next five years?

The next five years will shape who owns and benefits from AI across the Canadian economy. Many of those decisions are already being made through business contracts, public investments and workplace adoption.

For businesses, I’d prioritize retaining control of institutional knowledge and building internal capability. Before making a major AI investment, ask whether it will leave your organization more capable and less dependent over time.

For policymakers, I’d focus on turning the promise of shared prosperity into measurable outcomes. Canada’s AI for All strategy already recognizes that productivity gains should benefit workers, not just shareholders. The next step is to make that ambition visible in funding decisions, incentives and performance measures.

And for our institutions, I’d draw on something Canada already does well: shared and distributed ownership. Credit unions, co-operatives and dealer-owned retail businesses show that commercial success doesn’t have to depend on concentrating ownership in a few hands. We should bring that experience into the way AI businesses are financed, governed and developed.

Canada helped pioneer modern AI. We have an opportunity to do more than adopt what others build.

The question is whether Canadians will have a meaningful stake in the wealth this technology creates.

More from Retail Insider:

Mario Toneguzzi
Mario Toneguzzi
Mario Toneguzzi, based in Calgary, has more than 40 years experience as a daily newspaper writer, columnist, and editor. He worked for 35 years at the Calgary Herald covering sports, crime, politics, health, faith, city and breaking news, and business. He is the Co-Editor-in-Chief with Retail Insider in addition to working as a freelance writer and consultant in communications and media relations/training. Mario was named as a RETHINK Retail Top Retail Expert in 2024.

MORE FROM AUTHOR

Subscribe to the Newsletter

Subscribe

* indicates required

Related articles