Summary: AI Governance and Strategy in Canada

Session Overview

This executive summary covers a “Luncheon Session” from the CSPC 2025 conference featuring Mark Schaan, Associate Deputy Minister at Innovation, Science and Economic Development Canada, and Dr. Gail Murphy, Vice President of Research and Innovation at UBC. The discussion centered on Canada’s evolving Artificial Intelligence (AI) strategy, the transition from research leadership to industrial application, and the critical need for sovereign digital infrastructure and public trust.

Strategic Shift and Vision

While Canada was the first country to launch a national AI strategy in 2017 and has maintained research leadership for over 40 years, the government recognizes that “leadership is not a birthright”. The conversation highlighted a pivot from purely supporting research toward mastering commercialization, adoption, and productivity. A new, iterative strategy is expected in early 2025, informed by a recently completed “30-day national sprint” that engaged over 11,000 Canadians and an expert task force.

Key Priorities

• Sovereignty as Strategic Autonomy: Sovereignty is defined not as isolationism, but as possessing “strategic tradables”—assets valuable enough to ensure Canada has a seat at the table and can rely on domestic capabilities when necessary.
• Infrastructure Investment: The government is deploying over $2 billion in compute infrastructure, including subsidies for SMEs and a $700 million “Compute Challenge” to build domestic data center capacity for high-demand AI workloads.
• Trust and Safety: acknowledging that Canada has some of the lowest AI trust levels in the OECD, the government views building public confidence—through privacy modernization and safety institutes—as a prerequisite for economic adoption.

Detailed Conversation Summary

Legacy and Global Positioning

• Historical Leadership: Canada’s AI investment began 43 years ago with the Canadian Institute for Advanced Research (CIFAR), leading to a high concentration of Nobel and Turing Award winners.
• Global Competition: Competitor nations are making massive industrial investments, necessitating that Canada continuously “earn” its leadership position.

Developing the New AI Strategy

• Consultation Process: Rather than a year-long panel, the government executed a “30-day national sprint” in late 2024 to gather rapid feedback.
• Task Force: A 28-member task force submitted 31 reports to inform the strategy.
• Iterative Approach: The upcoming strategy (expected early 2025) will be a “living” document that evolves alongside the technology, rather than a static multi-year plan.

Sovereign Compute and Infrastructure

• Defining Sovereignty: Sovereignty is framed as “strategic autonomy”—having the option to rely on oneself and possessing assets (tradables) that partners value.
• SME Support: A $300 million program provided subsidized compute access to over 1,300 small and medium enterprises (SMEs), exceeding demand expectations by four times.
• Large-Scale Infrastructure: A $700 million “Compute Challenge” targets two groups: companies with massive AI workloads (e.g., large language models) and mid-tier data center operators expanding their sovereign offerings.
• Future Plans: A Sovereign Compute Infrastructure Program is being designed to support the research community and potentially utilize government tenancy to de-risk new data center builds.

Data Governance

• National Asset: Data is viewed as the “lifeblood” of the digital economy and the center of public confidence.
• Priorities: The focus is on structuring data for business processes, ensuring interoperability, and modernizing private sector privacy laws to allow data to be used safely.

Trust, Adoption, and Safety

• Trust Deficit: Canada ranks among the lowest in the OECD for trust in AI technology. Mark Schaan noted that “adoption moves at the speed of trust”.
• Labor Concerns: The government aims to counter fears of displacement by framing AI as an augmentation tool that elevates human work in the value chain.
• Safety Measures: Efforts include the Canadian AI Safety Institute, voluntary codes of conduct, and international collaboration on safety standards.
• Regulatory Philosophy: Regulation focuses on use cases and personal information protection rather than banning whole AI systems, though “no-go” zones may exist where privacy violations are inherent.

The “Triangle” of AI Adoption

• The Policy Challenge: It is easier to support the “top of the triangle” (AGI researchers) or the middle (AI-native companies) than the “bottom of the triangle”—mass adoption across 41 million Canadians and 400,000 businesses.

Goal: The ultimate goal is to bridge the gap between brilliant research and widespread economic application to solve Canada’s productivity and sovereignty challenges.