A new report from the Linux Foundation, in cooperation with Meta, has a stark message for Canada’s AI ambitions: despite taking a leading role in AI research, Canada risks falling behind in turning those same AI research strengths into a future economic advantage.

Founded in 2000, the Linux Foundation is a non-profit organization that serves as a neutral, trusted hub for fostering open-source software, hardware and standards. It is supported by over 1,000 corporate members, including major technology players like Google, IBM and Intel.

The study, The Value of Open Source AI for the Canadian Economy, put this bluntly: “Today, Canada continues to produce 10 per cent of the world’s leading AI researchers and ranks among the top five countries globally for highly cited research publications and newly funded AI companies. Yet Canada’s early AI investments in research and integration have not manifested into market advantage.”

This is not due to a lack of funding in AI development. Government investment in AI has remained strong with the 2024 federal budget containing a $2.4 billion AI infrastructure package, the study points out. Soon after, the government created an AI task force with a mandate to develop a national AI strategy.

While AI research in Canada continues to move at a strong pace, the study’s authors find that research has not translated into AI becoming commercialized or widely adopted by Canadian firms. The study finds only 26 per cent of Canadian organizations have rolled out AI in their firms, behind the 34 per cent of global firms taking on AI solutions.

A 2024 survey in partnership with the Vector Institute, and cited in the study, found 69 per cent of Canadian B2B businesses remain in the “crawl” or “walk” phases of adoption, compared with higher maturity levels internationally. Canadian firms also report lower adoption rates of AI pilots, and fewer efficiency gains, pointing to a slower pace of integration into business operations, the study notes.

Open source AI as Canada’s competitive advantage

To accelerate the move from research to commercialization, the Linux Foundation study argues for the wider adoption of open source AI — that is, models whose architecture and documentation are freely available.

Open source, the study’s authors say, “lowers barriers to entry, enhances transparency, and enables model fine‑tuning,” and would allow Canadian companies to innovate faster and at lower cost. Open source AI would also support Canada’s goal of digital sovereignty by assuring Canadian data stays in Canada and commercialized AI solutions can reflect Canada’s diverse linguistic heritage, identities and provincial and federal government regulatory environments.

The study’s authors argue that open source AI is being adopted by more startups, allowing them to move AI research and development to commercialization faster. This attracts more investment which will then fuel faster adoption of AI into companies and bring positive benefit to Canada’s economy.

“Research shows that AI could add up to nine per cent to Canada’s GDP by 2035 and $180 billion annually by 2030,” the study’s authors note. “Generative AI alone is projected to raise worker productivity by eight per cent and create over 35,000 innovation-driven jobs within five years.”

Several Canadian sectors are seeing the benefits of AI adoption. In manufacturing, firms like Magna and Bombardier have achieved 35 to 40 per cent reductions in maintenance costs through predictive AI and the government could unlock $14 billion annually.

“Canada now stands at a pivotal moment in shaping its AI future,” the study’s authors add. “If Canada acts decisively now by embedding openness into policy, practice, investment, and innovation, it can transform its early leadership on AI into a lasting global advantage.”