Artificial Intelligence Minister Kanishka Narayan has agreed a renewed AI security partnership between the United Kingdom and Canada.
The arrangement will deepen co-operation between the UK AI Security Institute and its Canadian counterpart as frontier models become more capable.
The two countries will share information about emerging risks and capabilities, including through appropriate Five Eyes channels. They will also exchange lessons on building domestic resilience to AI-related threats.
Technical work will include regular exchanges on evaluation methods and best practice. The governments will explore staff exchanges and co-ordinated model evaluations to strengthen links between the institutes.
The partnership also covers interoperable security standards, with the aim of making appropriate practices easier to recognise and adopt internationally.
The agreement builds on a science-of-AI-safety partnership signed in 2024 and a wider UK-Canada growth and innovation partnership from 2025. It does not announce a new funding allocation.
