Investing in dual-use, military-civilian AI compute infrastructure could help Canada meet its defence and security challenges by providing Canada’s military with a turn-key AI compute capacity for use during crises, while providing Canadian students, researchers, and innovators with reliable, low-cost AI compute access during peacetime.

The federal government is set to significantly increase its defence spending in the coming years to meet Canada’s NATO spending commitments and address an increasingly challenging global security situation. 

Canada’s move to strengthen national defence comes at a time when advancements in artificial intelligence (AI) technology are demanding significant investments in advanced computing capacity, including the construction of AI data centres. Leveraging AI is essential for supporting research and innovation in today’s economy. AI technology also has significant national security and defence applications.

Dual-use AI compute: Canada’s opportunity to support both national defence and civilian innovation

To increase access to AI compute for research and innovation, as well as security and national defence, the federal government should invest a portion of new defence spending into dual-use AI compute capacity—meaning compute capacity that has both military and civilian applications. 

During peacetime, this capacity could be provided to Canadian students, researchers, and innovators, helping Canada realize the scientific, technological, and commercial promises of AI through reliable and low-cost AI compute access—while during an emergency or international crisis, it could be streamed towards military and security applications, rapidly deploying reserve AI compute power to support the Canadian Armed Forces. Compute power could also be provided to Canada’s allies through security partnerships like NATO, or in support of international disaster response or UN Peacekeeping efforts.

Canada’s increase in defence spending is an opportunity for critical investments into AI

By the end of the year, Prime Minister Mark Carney has committed to raising Canada’s defence spending to 2% GDP, in line with NATO’s 2014 spending commitments. Furthermore, Canada has agreed to meet NATO’s new defence spending target of 5% GDP by 2035—split between 3.5% on core defence spending and 1.5% on infrastructure related to defence and security. 

According to forecasts by the Stockholm International Peace Research Institute, for Canada to meet the 5% GDP target, it would need to spend a combine US$170.3 billion (C$235 billion) on defence annually by 2035. Annual spending would include US$119 billion (C$164 billion) to meet the 3.5% core defence spending target, as well as an additional US$51.3 billion (C$70.8 billion) to meet the 1.5% target for supplementary defence infrastructure-related spending.   

To date, policy discourse around this additional 1.5% in defence infrastructure spending has been focused on its potential to fund needed critical infrastructure, such as seaports and airports. 

The Canadian military’s 2024 Artificial Intelligence Strategy envisions AI to have important national defence applications. Yet during a conflict or international crisis, AI compute capacity may be difficult or impossible to acquire. Secure, available, and scalable domestic AI compute capacity is needed to support future Canadian Armed Forces operations. Military AI use cases include autonomous systems, battlefield management and communications, intelligence and cybersecurity, simulation and training, logistics and predictive maintenance, and medical applications.

Dual-use, low cost, equitable access to AI compute for Canadians

There is a significant opportunity to allocate a portion of these investments into the development of domestic AI compute capacity to strengthen national defence, while underpinning Canadian digital transformation, sovereignty, and resilience. 

As dual-use digital infrastructure, a military-civilian AI compute capability could be used to provide low-cost and equitable access to AI computing power for students, researchers, and innovators. 

Accessible compute capacity could be put to work in college and university classrooms across Canada, allowing students to experiment directly with AI technology, building the next generation of AI technical talent. It could be used by Canadian researchers towards key national priorities, such as the green transition and sustainability, enhancing productivity, and increasing Canada’s cybersecurity posture and digital resilience.

It would help spur innovation, by allowing entrepreneurs access to low-cost compute capacity to develop new AI models and experiment with novel AI-powered products and services. Dual-use AI compute capacity could also be used as a catalyst to help Canada build international partnerships towards research, training, and responsible AI development.    

Canada must keep up with AI compute development in peer economies

Research by McKinsey & Company suggests that globally, data centres will require $6.7 trillion in new investment by 2030 to keep up with current computing demand, with $5.2 trillion of this investment required for AI processing tasks.

To meet this surge in demand, peer economies are developing significant AI compute infrastructure to meet national needs. In early 2025, the United States launched the Stargate Project, a four-year, public-private initiative to develop $500 billion in AI data centres across America. In July 2025, The White House released a national AI action plan, which included provisions to accelerate AI data centre development. The European Union launched its €200 billion InvestAI program in February 2025, which includes €20 billion to build a network of AI gigafactories in Europe. 

These announcements demonstrate the significant investments peer economies are making to support AI compute infrastructure, though Canada has also made recent strategic investments.

In late 2024, the federal government released its Canadian Sovereign AI Compute Strategy. The strategy earmarked $2 billion over five years towards building out national supercomputing capacity, providing a $300 million AI Compute Access Fund for innovators and businesses, along with ecosystem funding intended to catalyze private investment in domestic AI infrastructure. 

Even if a modest portion of the 1.5% of the supplemental defence infrastructure-related spending were earmarked towards building a Canadian military-civilian AI compute capacity, it would, over the next decade, greatly build upon investments already being made through Canada’s sovereign AI compute strategy.     

Canada is a medium-sized economy currently operating under significant fiscal pressures. Canadian policymakers need to be strategic and creative when it comes to earmarking new funding for national projects. Investing in dual-use AI compute capacity—which can be used by students, researchers, and innovators in peacetime and the military during a crisis—will help Canada foster its research and innovation ecosystem, increase its digital resilience, and protect its national security and sovereignty.

The analysis and opinions expressed in this article are those of the author and do not necessarily reflect the views and opinions of ICTC.