The UK and Canada have announced a renewed AI security partnership with a strengthened focus on sharing of information on frontier AI risks and capabilities, best practices in AI evaluation, exploring staff exchanges and coordinated evaluations.

The agreement builds on the UK-Canada Science of AI Safety Partnership first announced in May 2024, and the UK-Canada Growth and Innovation partnership signed in 2025.

The partnership will combine Canada’s AI research strengths with the UK’s expertise in AI testing and evaluation, supporting both countries’ efforts to understand and respond to emerging risks.

The renewed agreement builds on the two countries' national security partnership and strong existing collaboration to ensure safe and responsible adoption of AI around the world, It will focus on three key areas.

Firstly, action will be taken on AI security standards, including collaborating on the development of interoperable standards to enable recognition and broad uptake of appropriate AI security practices; and, identifying priority areas for future cooperation on common, international AI security standards.

Information-sharing agreements will be enhanced, including through appropriate Five Eyes channels for sensitive reporting, to support deeper collaboration on AI security challenges. The two countries will also share lessons on building domestic resilience to promote best practice and strengthen our respective societies and economies.

Finally the partnership agrees to expand the focus of technical cooperation, including establishing regular exchanges on frontier AI capabilities, emerging risks, evaluation methodologies and best practice; and, exploring staff exchanges and coordinated evaluation activities to strengthen institutional links and knowledge-sharing.