Canada keeps talking about sovereignty like it’s a set of hockey cards

“Sovereignty” is getting that treatment.

Right now, too much of the conversation treats sovereignty like a set of hockey cards. Slap the right logo on it. Make sure it looks Canadian from the front. Try not to look too closely at who actually owns the team, controls the rink, or decides when the lights go out.

That is the problem.

Canada has started using “sovereignty” to describe almost anything that sounds vaguely domestic. Put the data here. Add a local partner. Rent infrastructure from somewhere else. Keep the same dependency chain, the same exposure, and the same operational risk, but give it a Canadian accent and suddenly everyone starts speaking in solemn tones about national control.

That is not sovereignty. That is dependency in a Team Canada jersey.

What made the BetaKit piece useful is that it cut through some of that fog. The real issue is not whether the product looks Canadian in the brochure. The issue is whether Canada, or a Canadian organisation, actually controls the capability in any meaningful sense. Who holds the power? Who can compel access? Who controls operations? Who can interrupt service? Who owns the recovery path when something goes sideways?

That is where the conversation gets less fun for the people selling branding and more useful for the people responsible for risk.

Because sovereignty is not about optics. It is about control.

And control is where a lot of these nice tidy narratives start to wobble. If the core capability still depends on a foreign provider, if the underlying infrastructure can still be governed elsewhere, if the commercial terms can still be changed on you by someone with vastly more leverage, then the sovereignty story is doing a lot of work that the architecture is not.

We have become very good at confusing “hosted here” with “controlled here.”

Those are not the same thing.

You can host something in Canada and still be structurally dependent. You can put a domestic wrapper around a service and still leave the real power somewhere else. You can build a perfectly respectable local offering that, under pressure, still turns out to rely on someone else’s rules, someone else’s platform, and someone else’s tolerance for inconvenience.

That is why the “kill switch” question matters so much. Not because it is dramatic, although it is. It matters because it forces everyone to stop admiring the hockey card and start asking who owns the league.

If someone else can interrupt the service, revoke the capability, control the access path, or dictate your recovery options, then the sovereignty claim is, at best, incomplete. At worst, it is decorative.

That is exactly where SkyeConnex fits.

SkyeConnex was built around a fairly simple idea: if one provider holds the whole file, then one provider holds too much power. Too much operational power. Too much legal exposure. Too much leverage. Too much of your resilience story in their hands.

So instead of treating sovereignty like a region setting or a procurement adjective, we treat it like an engineering problem.

Reduce single-provider dependency.

Reduce the amount of control any one cloud has over the whole outcome.

Reduce the risk that one outage, one breach, one policy change, one legal order, or one strategic shift somewhere far above your pay grade becomes your emergency by lunchtime.

That is a much more serious definition of sovereignty than the one we usually get, which is essentially, “Don’t worry, it has a maple leaf on the slide.”

And that is really the point. Sovereignty should not be judged by the front of the hockey card. It should be judged by what is on the back: who owns it, who controls it, what the dependencies are, what the exposure is, and what happens when the game turns ugly.

Because that is when you find out whether you bought an actual capability or just a collectible.

Canada needs to get more precise about this, especially now, when sovereignty is becoming a fashionable word in everything from defense to cloud to AI to critical infrastructure. Precision matters because vague definitions are incredibly convenient for vendors, politicians, and anyone else who would prefer not to answer awkward architectural questions in public.

Once the word gets fuzzy enough, everyone qualifies.

And when everyone qualifies, the term becomes useless.

That is why I think this debate matters. Not as a branding argument. Not as a policy fashion trend. But as a practical test of whether Canada is serious about building systems it can actually rely on.

If the answer is yes, then we need to stop treating sovereignty like a collector’s item and start treating it like a control framework.

That means asking harder questions. Not “Is it Canadian-looking?” but “Who holds the power?” Not “Where is the rack?” but “Who can reconstruct the data?” Not “Is there a local logo involved?” but “Can one provider still control the outcome?”

That is where SkyeConnex belongs in the conversation.

Not as sovereignty theatre. Not as another sticker on the helmet.

As part of the actual architecture of reducing dependency, distributing control, and making sure no single provider gets to hold the whole file, the whole exposure, or the whole leverage.

Because if sovereignty is real, it has to survive contact with engineering.

Otherwise it is just a very nice hockey card.

Defense tech founder warns Canada is getting “wishy-washy” about defining sovereignty | BetaKit

Originally published by Ross Norrie, founder of SkyeConnex, on LinkedIn.

Published April 11, 2026 · More from the SkyeConnex blog