In February, Prime Minister Mark Carney launched Canada’s first defence industrial strategy, promising to lift the share of defence procurement awarded to Canadian companies to 70 per cent over the next decade from the current 43 per cent.

It is the most ambitious procurement target Ottawa has set in a generation. However, it has a major flaw. The strategy never defines what makes a company “Canadian.” That is not just an omission. It is the whole game.

Guillaume Beaumier and Hubert Cadieux argued recently in Policy Options that Canada risks “funding its own dependence at a cost of billions of dollars” by branding compute capacity built on American hyperscaler clouds as “sovereign.”

Their warning was about artificial-intelligence infrastructure. The same logic applies – with sharper edges – to defence software.

Canada’s digital sovereignty debate overlooks software risks The right way for Canada to secure cloud sovereignty The secret to a successful defence industrial strategy is procurement reform

Canada’s digital sovereignty debate overlooks software risks

The right way for Canada to secure cloud sovereignty

The secret to a successful defence industrial strategy is procurement reform

If we cannot say what “Canadian” means for procurement, the 70-per-cent target can be hit merely by relabelling the status quo. Instead, the federal government should implement three tests – covering auditability, data residency and Canadian-controlled intellectual property (IP) – before any software can be counted as Canadian for procurement purposes.

The argument that Beaumier and Cadieux made about compute and the argument about software lead to the same operational definition. Sovereignty is the effective control of the digital systems on which the........

© IRPP - Policy Options