The Weekly Brief — Issue 07

Hello,

The AI boom is colliding with something very physical: electricity.

Ontario is proposing new rules for large data centres seeking grid access, including a government approval process and a separate electricity rate class. At the same time, Alberta’s utility regulator is opening a hearing on a proposed 1,494-megawatt power plant designed largely to supply on-site AI and high-performance-computing data halls.

Plus: Ottawa launches a manufacturing workforce alliance shaped partly by AI adoption, and new Canadian research asks why promising firms sell just as they need to scale.

here is the week.

What happened

Ontario opened a 30-day consultation on August 13 on how large new data centres should connect to the province’s electricity system.

The province is considering requiring covered large new data centres to obtain government approval before connecting or reconnecting to the grid. It is also proposing a separate electricity rate class for new data centres above a threshold such as one megawatt, with the goal of keeping the costs and planning risks created by new demand from being shifted onto existing electricity customers.

Ontario’s proposed “Data Centre Playbook” would assess projects on economic development, digital sovereignty, community benefits, security and environmental considerations. The proposal also includes a preference for Canadian-owned and operated data centres as part of its digital-sovereignty pillar.

The scale behind the consultation is significant. Ontario’s Ministry of Energy and Mines says data-centre connection proposals could exceed 10,000 megawatts cumulatively. Recent provincial peak demand has been about 25,000 megawatts.

Why it matters

AI can feel weightless when it arrives through a browser or an app. The infrastructure behind it is not.

The Independent Electricity System Operator’s 2026 outlook expects data centres to account for 8.6 per cent of Ontario electricity demand by 2050 in its reference scenario, about 60 per cent more than its previous forecast. The IESO also treats data-centre growth as one of the more uncertain parts of its long-term demand outlook.

Ontario’s proposal turns that uncertainty into a practical policy question: when many large projects want access to the electricity system, should approval consider only reliability, technical feasibility and cost responsibility — or also the economic, sovereignty and community value a project brings?

The 10,000-megawatt figure is the cumulative size of connection proposals. It is not a forecast that Ontario will actually build or connect 10,000 megawatts of new data-centre demand. The approval process, threshold, exemptions and rate design are also still proposals under consultation.

Your move

If your organization is buying AI or cloud services, add one infrastructure question to the usual cost and privacy checks: where will the computing actually happen?

You do not need to become an energy expert. But knowing where data is hosted, which jurisdiction governs it and how dependent a service is on scarce infrastructure is becoming part of understanding the real risk and resilience behind AI.

Alberta opens a hearing on a power plant for an AI campus

The Alberta Utilities Commission is scheduled to begin a hearing Tuesday on a proposed 1,494-megawatt natural-gas power plant near Indus, east of Calgary. Under normal operating scenarios, about 1,200 megawatts would supply co-located data halls intended for artificial-intelligence and high-performance-computing facilities.

The project is still under review. A previous federal decision that no further assessment was required under the Impact Assessment Act is not the same as provincial approval, and the AUC hearing has not yet produced an outcome.

Why it matters: Ontario is considering how very large data centres should draw from a shared provincial grid. The Alberta project illustrates another path: pairing a large computing campus with generation built largely to serve it. Either way, AI infrastructure is becoming an energy-planning issue, not just a technology story.

Ottawa launches a manufacturing workforce alliance

The federal government launched the Advanced Manufacturing Workforce Alliance on Monday, bringing together manufacturers, unions, education and training institutions, industry groups and other partners to identify workforce gaps and develop practical training responses.

Ottawa says advanced manufacturing is increasingly shaped by automation, robotics and AI-enabled systems. The alliance is the second of six sector-focused Workforce Alliances. The broader program has an $81-million, five-year envelope across all six sectors; that money is not dedicated solely to manufacturing or AI. Canada’s manufacturing sector employed about 1.85 million people in 2024.

Why it matters: This is an early sign of Ottawa treating AI adoption as a workforce-transition problem as well as a technology-investment problem. The important test comes next: whether the alliance produces training that workers and employers can actually use, rather than another layer of coordination.

Canada has a scale-up gap, CCI says

The Council of Canadian Innovators released a new qualitative study based on interviews with 30 founders whose Canadian companies were acquired by foreign buyers. CCI says recurring barriers included access to growth capital, early domestic customers, fragmented support, experienced talent and infrastructure.

The study is not specific to AI companies, and CCI is an advocacy organization, so its findings should not be treated as a census of the Canadian technology sector. But the timing is relevant: Canada’s national AI strategy says it wants to help Canadian AI companies scale by unlocking growth capital and using government as a strategic anchor customer.

Why it matters: The two documents point at the same basic test from different directions. Ottawa says capital and government customers can help Canadian AI firms scale here. CCI’s founder interviews suggest those are exactly the kinds of gaps that can push promising companies toward a sale. The question is whether the new AI strategy changes that pattern in practice.

A megawatt measures power demand at a point in time. When Ontario says proposed data-centre connections could total more than 10,000 megawatts, it is describing the combined size of projects that have asked for connections — not electricity that has already been committed.

That distinction matters. Applications give grid planners a signal about possible future demand, but projects can change, shrink, be delayed or never connect at all.

The useful takeaway is not “Ontario needs 10,000 megawatts tomorrow.” It is that the queue is large enough for the province to decide it needs new rules for how major data-centre loads are assessed and charged.

Pick one AI tool you use at work and answer three questions:

If you cannot answer one of them, that is the useful finding. Do not enter confidential, personal or sensitive information into a public AI tool to test this.

A calm, practical briefing on the AI developments that matter in Canada—delivered every Tuesday.

That’s the week. See you next Tuesday.

— Padge T.

AI Brief Canada is an editorially independent publication published by Groundshift Advisory Inc. and written by Padge T.

Groundshift Advisory helps Canadian organizations adopt AI responsibly. Sponsored, affiliated, or commercially connected content will always be clearly disclosed.

Issue disclosure: Nothing in this issue is sponsored, affiliated, paid, or commercially connected.

AI use: AI tools supported research, source discovery, outlining and drafting for this issue. All material was reviewed, edited and fact-checked by Padge T. before publication.

Not affiliated with the Government of Canada.

Corrections or feedback: We welcome corrections and reader feedback. Email [email protected] or review our Editorial Policy.