Skip to content

News

Abstract graphic
News

Response to Canada’s Sovereign AI Compute Strategy

Dec 5, 2024

Today, the federal government announced a $2-billion investment in Sovereign AI Compute Strategy, the majority of the investment from the 2024 budget’s $2.4-billion Securing Canada’s AI Advantage. The detailed plan of the AI Sovereign Compute Strategy focuses on three objectives: 

  1. $700-million investment in new or expanded data centres through the AI Compute Challenge, improving domestic infrastructure for commercial adoption
  2. $1-billion to build AI-specific public computing infrastructure, accessible by Canadian researchers and innovators
  3. $300-million through the AI Compute Access Fund for small- and medium-sized enterprises (SMEs) to reduce cost-related adoption barriers

The three funding components outlined in today’s announcement align in many ways with what we proposed in our September 2024 submission to the Government of Canada’s consultation on AI Compute. In particular, we called for commercial data centre expansion, public infrastructure modernization, and immediate compute access for small- to medium-sized enterprises, which have been reflected in the new strategy.

AI Computing power—the machinery and software layers that allow others to develop or use artificial intelligence applications—is too expensive and out of reach for too many Canadian companies, researchers and governments. Our March 2024 report, “Can Canada Compute?” showed how far behind Canada is in developing this crucial part of the AI economy. Today’s announcement won’t bridge the gap alone, but it demonstrates that the government’s commitment is ongoing.

We’re especially intrigued by the specific commitment around data centres, which suggests the Government believes that Canada has a comparative advantage in developing this key building block for the world’s AI infrastructure.

But data centres aren’t easy to build. And there’s more work to be done. 

As the government moves forward to develop program details, we urge Canada’s leaders to consider: 

  1. The intersection of energy and climate impacts with regional economic benefits of infrastructure expansion. Expansion must have a lasting long-term impact on living standards for affected communities. Proposed projects and partnerships must consider the energy efficiency and environmental impacts of data centre expansion in accordance with the OECD’s framework on AI compute and energy.
  2. The economic and social impacts of public AI infrastructure. Canada must prioritize AI innovation and research that fosters public access, accountability, and welfare. We must ensure IP generated translates to regional economic development, business dynamism, and responsible use and adoption of the technology. For example, the proposed evaluation criteria for infrastructure expansion must not crowd out domestic AI supply chain firms that provide long-term economic value and high-skill job creation (e.g., telecommunication infrastructure, and semiconductor design firms) for Canada’s economy. 
  3. The need for an expert intermediary organization to evaluate the long-term economic impacts of infrastructure expansion, commercial value creation, and technical requirements of Canada’s AI ecosystem proposals. The entity, closely engaged with public and private stakeholders, would provide the economic, commercial, and technical expertise to ensure that public investments translate to a maximum return to Canadian society and the economy.

The government also outlined the findings of its recent consultations on the strategy, which largely reflect the challenges and recommendations that have emerged from roundtables conducted by the Dais with AI industry and ecosystem players (published recently in our report, From Potential to Performance). We are supportive of the overall thrust of other submissions, highlighting the need for better data stewardship and access for innovation and research, as well as addressing the critical gaps in training for AI adoption and commercialization efforts in Canada.

The Dais is committed to delivering swift, actionable insights on AI infrastructure, talent, regulation, and adoption as these programs evolve. From exploring the link between sustainable energy and data centre development, to empowering underserved communities to understand and benefit from AI, we are eager to drive the responsible use of this transformative technology.

—

Connect with Dais experts for comment on this announcement, and the latest research on AI infrastructure:

Experts in AI Compute

Graham Dobbs | Senior Economist (graham.dobbs@torontomu.ca) 
Jake Hirsch-Allen | Senior Fellow (jake.ha@torontomu.ca) 

Media Inquiries

Nina Rafeek Dow | Communications and Marketing Lead (nina.rafeek@torontomu.ca)