Policy Framework for the Future of Artificial Intelligence in the 21st Century

VIEW / DOWNLOAD OUR FULL SUBMISSION

Executive Summary

1. This report presents policy recommendations for the Canadian federal government based on a roundtable discussion convened by the Internet Society, updated to reflect federal policy announcements through October 2025. The roundtable brought together experts in artificial intelligence, digital policy, cybersecurity, and innovation to address Canada’s strategic positioning.
2. The roundtable identified a critical juncture for Canada between complete systemic failure and complete societal transformation rather than incremental policy adjustments. Since the discussion, the federal government has taken significant steps, including Minister Evan Solomon’s announcement of a 26-member AI Strategy Task Force conducting a 30-day national consultation, with an accelerated timeline promising to table the refreshed strategy by year-end 2025.
3. Central conclusions of the roundtable that are consonant with recent government policy include implementing immediate governance practices, developing domestic foundational AI technologies to reduce US dependency, and emphasizing digital sovereignty as “the most pressing policy and democratic issue of our time.” The government’s $2 billion Sovereign AI Compute Strategy directly addresses technological dependency concerns raised in the roundtable. Recent stakeholder consultations, including the October 22, 2025, TCIS roundtable, have reinforced that trust is not a barrier to innovation but a prerequisite for it, and that sovereignty must be understood as ‘agency’—the ability of individuals, corporations, and government to control their digital futures rather than merely data residency.
4. Critical gaps remain, including comprehensive workforce retraining programs for AI displacement, a framework for data governance balancing individual privacy and consumer protection with corporate use and public interest, aggressive immigration policies to attract talent, and procurement reform prioritizing Canadian AI companies over established international vendors.

Background and Context

1. Canada stands at a pivotal moment in AI policy development that has gained unprecedented momentum since the June 2025 roundtable discussion. As the first country to launch a national AI strategy in 2017, Canada established early leadership, yet the global landscape has transformed dramatically, requiring fundamental reassessment. Minister Evan Solomon’s appointment as Canada’s first Minister of Artificial Intelligence and Digital Innovation brought renewed focus and an accelerated timeline.
2. The September 2025 ALL IN conference in Montreal marked a crucial inflection point, with Minister Solomon announcing the AI Strategy Task Force and emphasizing digital sovereignty as a core priority. The 26-member task force includes prominent figures identified in the roundtable discussion, though it lacks the diversity one would hope to see in such an important policy initiative. Minister Solomon’s commitment to modernizing “Canada’s privacy and data laws [that] are more than 25 years old” somewhat addresses concerns about outdated regulatory frameworks, but with a narrow focus on privacy that does not address new issues raised by the technology, such as accountable AI.
3. The international context continues to validate concerns about Canadian technological dependency. Recent Microsoft statements that US law takes precedence over Canadian data sovereignty requirements have reinforced the urgency of building domestic AI infrastructure. The current policy environment reflects critical tensions between regulatory protection and innovation competitiveness, waiting for comprehensive legislation versus implementing immediate governance practices, and positioning Canada strategically while maintaining sovereign control. The October 2025 AWS outage affecting Canadian users provided concrete evidence of foreign infrastructure dependency risks.
4. This context is complicated by Canada’s unique structural challenges, distinguishing it from other G7 economies. Unlike the United States, where 50% of private sector employment occurs with large companies capable of significant AI investments, 90% of Canadian employment occurs in small and medium enterprises with limited resources for AI adoption, requiring tailored policy approaches recognizing Canada’s distinct innovation ecosystem.

Problem Statement

1. Canada faces interconnected challenges threatening its position as a global AI leader and economic competitiveness in the digital age, with recent government actions partially addressing some concerns, while others remain pressing. The core problem manifests across three critical dimensions that have been partially but not comprehensively addressed by federal initiatives.
2. Regulatory uncertainty continues despite progress, as the absence of comprehensive AI legislation creates ongoing uncertainty for Canadian businesses, even with the AI Strategy Task Force’s establishment. Industry stakeholders report reluctance to make major AI investments without clear regulatory frameworks, while the complexity of existing privacy and consumer protection laws creates compliance challenges when applied to AI systems. The stalled AIDA legislation left Canada without risk-based governance frameworks for high-impact AI systems. Recent consultations reveal Canada lacks a coherent regulatory framework across overlapping areas, including privacy, cybersecurity, AI regulation, competition, and online harms. Practitioners emphasize differentiating what should be in legislation versus regulation versus standards versus voluntary codes to avoid box-ticking rules burdening SMEs.
3. Technological dependency and sovereignty risks remain significant despite targeted government investments. Canadian AI companies demonstrate dangerous dependency on US-based Large Language Models and cloud infrastructure. The roundtable identified this as an existential threat, noting that if US providers restrict access, Canadian businesses could “collapse like dominoes.” This dependency extends beyond software to fundamental infrastructure, with American firms owning significant portions of Canadian data centers.
4. Economic transformation and skills displacement represent the most inadequately addressed challenge. The roundtable identified massive impending workforce disruption with AI poised to eliminate positions across sectors, including junior legal roles, cybersecurity analysts, and truck drivers. Despite corporate claims of augmentation rather than replacement, participants acknowledged the reality of job displacement, yet specific, comprehensive workforce retraining programs have not been announced.
5. These problems are exacerbated by cultural and institutional barriers remaining largely unaddressed, including risk-averse government culture prioritizing avoiding failure over achieving breakthrough innovations, procurement systems favoring established international vendors over Canadian startups, and interprovincial data-sharing barriers preventing leveraging Canada’s significant public data assets for AI innovation. Practitioners report procurement barriers, including coordination gaps between economic development and procurement departments, and repeated security vetting requirements rather than reusable certifications.

Analysis

1. Canada’s existing AI governance landscape reveals significant gaps and fragmented approaches that recent government actions have begun to address, while leaving substantial work remaining. At the federal level, the absence of AIDA has forced continued reliance on existing privacy laws and voluntary frameworks not designed for AI-specific challenges, though the AI Strategy Task Force’s establishment suggests movement toward comprehensive frameworks.
2. The government’s infrastructure initiatives show recognition of sovereignty challenges while requiring continued development to address the full scope of dependencies. The $2 billion Sovereign AI Compute Strategy represents the largest commitment to domestic AI infrastructure in Canadian history, directly responding to roundtable warnings about technological dependency. However, analysis reveals continued complexity in achieving true sovereignty, as the Cohere partnership, receiving $240 million, will operate from a data center with mixed ownership, including American firms.
3. Recent government actions show mixed progress on innovation ecosystem challenges identified in the roundtable. The emphasis on “accelerating adoption across the economy” addresses the invention-to-innovation gap, but specific mechanisms remain underdeveloped. The AI Strategy Task Force includes strong industry representation, suggesting recognition of the need for private sector engagement, yet procurement reform to support Canadian startups over established international vendors has not been explicitly addressed. Industry stakeholders emphasize that procurement represents industrial policy shaping the domestic AI ecosystem. Canadian sovereign cloud initiatives led by Evertz, Renovis, and ThinkOn demonstrate private sector commitment to building domestic infrastructure.
4. International positioning shows alignment with roundtable themes while requiring continued development. The government’s emphasis on projecting “Canadian values” through AI development reflects the trustworthy AI leadership concept discussed. However, specific partnerships with like-minded countries beyond general statements have not been announced, while the relationship with US systems remains complex.
5. Workforce transformation analysis suggests unprecedented disruption approaching that has received inadequate attention in recent government actions. Unlike previous technological transitions primarily affecting manufacturing, AI threatens knowledge work across sectors. The scale and speed of potential displacement exceed historical precedents and current retraining program capacity.

Policy Recommendations

1. Canada should immediately implement risk-based governance practices without waiting for legislative completion, requiring organizations deploying high-impact AI systems to conduct algorithmic impact assessments, implement bias mitigation measures, and establish incident reporting protocols. The federal government should prioritize the passage of a modernized AI and data act, incorporating stakeholder feedback from the previous AIDA process, defining high-impact systems clearly, and establishing proportionate requirements based on risk levels. Governance frameworks must ensure algorithmic fairness and inclusivity are integrated across design, deployment, and user experience phases.
2. Canada should implement a framework for data governance that adds public interest as a third pillar alongside individual privacy rights and corporate data use, following the taxation model where a portion of income serves the public good. This framework would require organizations to contribute data for public benefit projects under strict privacy protections, enabling innovation in healthcare and transportation while maintaining individual privacy rights.
3. Strategic AI infrastructure development must accelerate beyond current government commitments, requiring direct government investment in Canadian companies like Cohere, combined with policies encouraging domestic development of critical AI infrastructure. The approach should prioritize “intelligent interdependence” by accepting necessary foreign partnerships while maintaining control over critical capabilities. The federal government should accelerate development of sovereign cloud capabilities, ensuring Canadian data remains subject to Canadian law. Infrastructure planning must address technical architecture, including data architecture, MLOps processes, and model lifecycle management. Priority workloads for sovereign AI should focus on healthcare, defense, and critical public services.
4. Procurement reform represents immediate, actionable change requiring transformation of government practices to actively support Canadian AI companies rather than defaulting to established international vendors. Implementation should include risk assessment frameworks that properly evaluate Canadian startups alongside established providers, create set-aside programs for small and medium enterprises, and establish clear criteria for evaluating AI solutions that include sovereignty considerations. Procurement should be viewed as industrial policy where every decision strengthens or weakens the domestic ecosystem.
5. An aggressive immigration and talent retention strategy requires the development of fast-track immigration pathways for AI researchers, entrepreneurs, and technical professionals, particularly those fleeing unfavorable conditions in other jurisdictions. This should include streamlined work permits, accelerated pathways to permanent residence and citizenship, and targeted programs for entrepreneurs bringing job-creating opportunities, positioning Canada as the preferred destination for AI talent globally.
6. Comprehensive workforce retraining programs must be established immediately to address the pace and scale of AI-driven job displacement. These programs should focus on skills that complement rather than compete with AI capabilities, including human-centered design, complex problem-solving, and AI system oversight, designed for working adults and integrated with employment insurance to provide income support during retraining.
7. Implementation of AI dividend taxation for social support should capture value from AI productivity gains to fund comprehensive social support and retraining programs, including specific taxes on AI-generated productivity improvements or broader policies ensuring companies benefiting from AI adoption contribute to societal transition costs. Revenues should fund universal basic income pilots, expanded training programs, and community support initiatives.
8. Strategic international AI partnerships should be developed with like-minded countries, particularly in Europe, and with allies like Israel, to create alternatives to US-dominated AI systems. These partnerships should focus on joint development of AI technologies, shared research initiatives, and coordinated approaches to AI governance that balance innovation with the protection of democratic values.
9. Whole-of-government AI coordination must be established with coordinated strategies across all federal departments and agencies, creating clear ministerial responsibilities and regular progress reporting. This should include dedicated technical expertise in key departments, coordination mechanisms to prevent policy conflicts, and regular strategy updates to accommodate technological evolution. The government has established eight priority pillars: infrastructure, trust, security, research and talent, adoption, commercialization, scaling champions, and skills/education. Coordination must ensure these reinforce rather than duplicate each other.
10. Education system modernization at all levels must prepare students for an AI-integrated economy through digital literacy, AI ethics, human-AI collaboration skills, and adaptability training. Professional programs should integrate AI tools training while maintaining focus on distinctly human capabilities that complement artificial intelligence.
11. Public sector AI adoption leadership requires the government to lead by example in AI adoption, creating markets for Canadian solutions while demonstrating responsible use practices. This includes modernizing internal processes, providing data access for AI development under appropriate privacy protections, leveraging commitments to defence and other spending to include funding of AI projects, and showcasing successful Canadian AI implementations to international markets. The federal government has announced an AI Strategy for the Federal Public Service, demonstrating commitment to leading by example with transparency and auditability requirements.

Conclusion

1. Recent official policy statements and initiatives accord with our concerns and priorities and recognize the urgency and scope of transformation required. The establishment of the AI Strategy Task Force, acceleration of the AI strategy timeline from 2026 to 2025, and commitment to digital sovereignty acknowledge that Canada faces a moment requiring comprehensive societal transformation rather than incremental policy adjustments.
2. However, significant implementation challenges remain in areas requiring continued comprehensive attention, particularly workforce retraining for AI displacement, procurement reform to support Canadian startups over established international vendors, and interprovincial coordination to remove data sharing barriers for innovation. The government’s focus on building the “strongest economy in the G7” acknowledges the scale of ambition needed, but specific programs to achieve this transformation require continued development beyond current announcements.
3. The international context continues to validate the sovereignty concerns raised in the roundtable discussion, as Microsoft’s explicit statement that US law supersedes Canadian sovereignty requirements demonstrates the risks of continued dependence on foreign AI infrastructure. This reinforces the importance of domestic infrastructure investments while highlighting the complexity of achieving true technological sovereignty in an interconnected global economy. The choice between complete systemic failure or complete societal transformation remains before Canada, with recent government actions suggesting recognition of this reality and commitment to transformation over incremental approaches.

VIEW / DOWNLOAD OUR FULL SUBMISSION

 

References

Arnold, Brent, et al. “TCIS AI Roundtable Group 1 Meeting Transcript.” The Canadian Internet Society, June 24, 2025.

“TCIS AI Roundtable Summary.” The Canadian Internet Society, June 24, 2025.

Canada, Innovation, Science, and Economic Development. “Government of Canada launches AI Strategy Task Force and public engagement on the development of the next AI strategy.” September 26, 2025. https://www.canada.ca/en/innovation-science-economic-development/news/2025/09/government-of-canada-launches-ai-strategy-task-force-and-public-engagement-on-the-development-of-the-next-ai-strategy.html

Solomon, Evan. “Ministers Joly and Solomon meet the future of AI at ALL IN 2025.” Innovation, Science and Economic Development Canada, September 25, 2025. https://www.canada.ca/en/innovation-science-economic-development/news/2025/09/ministers-joly-and-solomon-meet-the-future-of-ai-at-all-in-2025.html

“Minister Evan Solomon reveals Canada’s AI Task Force.” BetaKit, September 30, 2025. https://betakit.com/minister-evan-solomon-reveals-canadas-ai-task-force/

“Canada hopes to build a sovereign cloud to counter US dominance.” BetaKit, September 28, 2025. https://betakit.com/canadian-sovereign-cloud-evan-solomon-all-in/

“Microsoft says U.S. law takes precedence over Canadian data sovereignty.” Digital Journal, August 15, 2025. https://www.digitaljournal.com/tech-science/microsoft-says-u-s-law-takes-precedence-over-canadian-data-sovereignty/article

Owen, Taylor. “A Canadian blueprint for trustworthy AI governance.” Policy Options, August 29, 2025. https://policyoptions.irpp.org/2025/07/canadian-ai-governance/

The Honourable Evan Solomon, Minister of Artificial Intelligence and Digital Innovation. Government of Canada. https://www.canada.ca/en/government/ministers/evan-solomon.html

Future Skills Centre. “Wendy Cukier.” https://fsc-ccf.ca/team/wendy-cukier-2/