Canada should own the stack: energy, data, inference, and Bitcoin

We export power and import permission

For decades, tight Canada–U.S. integration looked like pure efficiency. Sell resources into the larger market. Buy the finished service from the firms with the greatest scale. Let Microsoft, Google, Amazon, and other platforms absorb the capital cost of digital infrastructure. That bargain delivered real value.

It also concentrated control. Canadian organizations increasingly rent their storage, identity, collaboration, analytics, and AI from a small set of foreign-controlled platforms. Data may sit in a Canadian region while ownership, administration, telemetry, support, contracts, and legal exposure extend beyond Canada. A Canadian postal code is useful; it is not the same thing as Canadian operational sovereignty.

The tariff crisis makes the strategic weakness impossible to ignore. The lesson is not that every American company is hostile or that every cross-border service will be cut off. The lesson is that a foreign dependency remains foreign even when yesterday’s contract was friendly. Canada needs alternatives before a crisis, because alternatives assembled during a crisis are expensive and incomplete.

One sovereignty principle, four layers

These layers reinforce each other. Domestic energy makes compute possible. Domestic compute makes controlled inference possible. Controlled inference keeps Canadian knowledge and productive workflows available. Bitcoin provides a monetary rail outside the control of any foreign platform or single state. None replaces the Canadian dollar, the power grid, public institutions, or international trade. Together they create optionality.

Bitcoin is insurance, not a promise of price stability

Bitcoin is volatile. It has no central issuer promising redemption, and its market price can fall sharply. Calling it national or personal insurance does not make it a stable cash reserve, a guaranteed return, or a substitute for prudent fiscal and monetary institutions.

Its insurance property is different: a capped issuance schedule, global settlement, portability, and the ability to hold an asset without relying on a foreign bank, cloud account, or government bond issuer. At a time of rising sovereign debt, trade fragmentation, payment-platform concentration, and questions about the reliability of U.S. policy, that independent option is logically valuable.

The Bank of Canada’s own 2026 financial-system survey reports concern about rising U.S. debt, policy uncertainty, declining U.S. exceptionalism, foreign payment infrastructure, and foreign-controlled stablecoins. The Bank does not endorse Bitcoin as a national reserve, and D-Central should not pretend it does. The official concerns nevertheless support the broader conclusion: payment and monetary concentration are sovereignty risks worth diversifying.

Mining keeps part of the monetary infrastructure here

Canadians can buy Bitcoin mined elsewhere. But domestic mining develops a different capability: energy-to-Bitcoin conversion, pool and node operations, firmware, repair, power electronics, thermal engineering, and participation in the physical security of the network.

Mining should compete honestly for power. It should not displace households or productive industry at subsidized rates. Its strongest role is as a location-flexible and interruptible buyer near generation: absorb power when it is abundant, curtail when the grid is tight, reuse heat when an economic thermal customer exists, and pay for the infrastructure it causes.

Canada should also resist concentration inside mining. Foreign-controlled mega-farms running opaque firmware do little for personal or national sovereignty. A healthier ecosystem includes Canadian operators, transparent pools, independent nodes, open firmware, repair capability, smaller distributed miners, and clear rules that do not reserve participation for the largest balance sheet.

AI sovereignty means the right to run the model

Canada does not need every model to be invented here. It needs the practical right and capability to run suitable models here. That requires lawful access to model weights, compatible hardware, secure facilities, electricity, cooling, fibre, orchestration software, evaluation, and people who can keep the service alive.

Open weights do not automatically mean open source, safe, accurate, bilingual, or commercially licensed. Canadian operators must evaluate licences and task performance. But portable models make sovereignty possible in a way an API-only product cannot: if a provider changes price, policy, availability, or terms, the organization still has a path it can operate.

We should build Canadian cloud and colocation capacity without recreating a smaller monopoly. The goal is a competitive federation of regional operators, public and private compute, university and research capacity, on-premises systems, and interoperable software. Canadian firms should be able to move workloads among them.

The sovereign stack is not one Canadian hyperscaler. It is a country full of people and organizations capable of operating without a foreign off-switch.

Reducing two sides of the U.S. trade relationship

The Trump administration says it wants to reduce the U.S. trade deficit. Its tariffs are a destructive and often incoherent way to pursue that objective, but Canada can take the stated preference seriously on its own terms.

Use more Canadian electricity domestically and exports to the United States fall. Build Canadian data centres and inference services and imports of U.S.-controlled digital services fall. The resulting Canadian investment is not a tariff or a boycott. It is domestic capacity responding to revealed political risk.

The move also improves Canada’s trade mix. Instead of selling a larger share of a raw input and renting back the high-margin service, Canada can sell inference, hosting, software, security, and engineering while earning Bitcoin mining revenue over a global network. Digital services can be exported without the transmission losses and cross-border grid capacity required for electricity exports, while the facility, skilled work, and tax base remain here.

A Canadian sovereignty compact

1. Power Canadians first. Protect reliability and affordability, then evaluate domestic value-added uses before new long-term exports.
2. Build at the energy source. Zone and pre-engineer suitable compute sites near usable generation, with transparent interconnection and flexible-load tariffs.
3. Buy Canadian control. Public procurement should score residency, ownership, administrators, encryption keys, portability, Canadian support, and exit—not just the cloud region label.
4. Fund capability, not logos. Tie incentives to incremental generation, Canadian customer access, workforce training, open interfaces, cybersecurity, and auditable public benefit.
5. Protect permissionless participation. Let Canadians run nodes, self-custody Bitcoin, mine under fair energy rules, and use lawful open software without needing a dominant intermediary.
6. Test the exit annually. Every critical public and enterprise system should prove it can export data, replace a model, restore from offline records, and operate through a provider disruption.

D-Central’s role

D-Central comes from Bitcoin mining, where power, heat, firmware, networking, repair, and self-custody are physical realities rather than policy abstractions. The same operating culture belongs in Canadian AI: own what matters, understand the stack, prefer open interfaces, repair hardware, measure claims, and keep an exit.

We are building toward Canadian-controlled local and hosted inference while continuing to support decentralized Bitcoin infrastructure. We do not claim that one company can supply a national stack, that every workload belongs on-premises, or that every megawatt should mine. The point is to create enough capable Canadian operators that no single company—including D-Central—becomes the new permission layer.

Frequently asked questions

Does sovereignty require abandoning Microsoft, Google, or Amazon?

No. It requires understanding the dependency and maintaining a credible alternative. A governed hybrid system can use foreign platforms while keeping sensitive or continuity-critical workloads on Canadian-controlled infrastructure.

Should Canada hold Bitcoin as a national reserve?

That is a separate public-policy decision requiring legislation, custody design, governance, accounting, and risk limits. This article argues for preserving Canadian Bitcoin capability and treating the asset as one form of optionality, not for an unbounded treasury purchase.

Would domestic data centres eliminate foreign legal exposure?

No. Location, ownership, control, contracts, administrators, subprocessors, and applicable laws all matter. Canadian hosting can improve control, but buyers must verify the full operating and legal boundary.

Primary sources reviewed August 26, 2026: Bank of Canada, Financial System Survey 2026; Bank of Canada, financial markets assessment; ISED, enabling sovereign AI data centres; ISED, AI Compute Challenge; Office of the Privacy Commissioner, AI and your business; Canada Energy Regulator, electricity trade summary.