Can Canada ever have true digital sovereignty?
By now, Canada has had enough discussions about digital sovereignty to build a very impressive stack of panels, roundtables, and policy decks. Actual sovereignty is a bit harder.
We still tend to talk about it as though it is mainly a location problem. Keep the data in Canada. Use a Canadian region. Put a flag on the slide. Everyone relaxes for a moment and pretends jurisdiction, control, and technical dependency have all been dealt with by postal code.
They have not.
Joshua van Es has been pressing on exactly this point: sovereignty is not just about where infrastructure sits, but who owns it, who governs it, who operates it, and who can be compelled when things get serious. That is the real issue, and it is the part that gets politely skipped when the conversation becomes too convenient. (Open Canada)
That is also why I built SkyeConnex the way I did.
SkyeConnex was architected from the ground up around the idea that sovereignty is a control problem, not a branding exercise. Not “which cloud region would you like?” Not “which foreign hyperscaler would you prefer to depend on?” The design assumption was simpler and more useful: no single provider should ever hold enough of the data, keys, or control plane to reconstruct meaningful access on its own.
That changes the conversation.
Because if your data is merely sitting in Canada but remains fully exposed to the operational, legal, or technical reach of a single dominant provider, you do not really have sovereignty. You have residency with better marketing.
SkyeConnex was designed specifically to reduce that exposure. Split across providers. Encrypted end to end. No single cloud with the whole picture. No single provider holding the complete asset. No one vendor in a position to assemble, inspect, or surrender everything because, architecturally, they never had everything in the first place.
That is a very different model from the one most of the market still sells, which usually boils down to: “Trust us, but in a local region.”
Helpful, perhaps. Sovereign, not necessarily.
Canada can absolutely move closer to real digital sovereignty, but only if it stops confusing domestic hosting with domestic control. Real sovereignty requires Canadian governance where it matters, architectural separation where it counts, and less willingness to outsource critical digital dependency and call it strategy.
In other words, sovereignty has to be designed into the system. It does not appear because someone added a maple leaf to the procurement deck and said “trusted” three times.
That is the broader national issue.
The narrower and more practical one is this: if a country or enterprise wants real control, then its systems should be built so that no single outside actor can unilaterally reconstruct, expose, or surrender what matters most.
That was the thinking behind SkyeConnex from day one.
Source: Joshua van Es, Canada’s Sovereign AI Compute Gap: Why We Are Still Treating a Strategic Asset as a Service https://opencanada.org/canadas-sovereign-ai-compute-gap-why-were-still-treating-a-strategic-asset-as-a-service/ (Open Canada)
Originally published by Ross Norrie, founder of SkyeConnex, on LinkedIn.
Published March 25, 2026 · More from the SkyeConnex blog
More from the blog
Europe Isn’t Banning Microsoft. It’s Finally Asking the Right Question.
Useful Until It Isn’t
Everyone wants utility.
Read → Commentary · 2 min readData governance isn't failing because it's too slow. It's failing because we keep building the roof before the foundation.
Every few years, the data industry discovers the same problem wearing a new outfit.
Read → Commentary · 2 min readThe Government of Canada just published an AI Readiness Scorecard. Most departments should read it twice — and be honest the second time.
The Canada School of Public Service recently released a job aid to help public servants assess whether a problem is a…
Read →