The United Kingdom and Canada have pledged to strengthen collaboration on Artificial Intelligence (AI) safety, building upon their shared commitment to responsible AI development.

The announcement, confirmed by UK Technology Minister Michelle Donelan and Canadian Science and Innovation Minister François-Philippe Champagne, establishes a new partnership between the two countries’ respective AI Safety Institutes (AISIs).

Donelan tweeted “The collaboration of AI Safety Institutes is vital to the responsible development of AI and the advancement of science worldwide.”

The AI Safety Institute added, “This partnership is a key moment in the development of an international network of AI safety institutes.”

Both governments recognise the importance of responsible AI development to harness its immense potential. This partnership sparks an open global dialogue, drives foundational research, and promotes international alignment on AI safety best practices.

By joining forces, the UK and Canada are not only accelerating the development of robust AI safeguards, but also setting a global example for collaborative and responsible AI advancement.

Key Aspects of the Partnership:

• Sharing expertise: Both countries will exchange knowledge and best practices to strengthen AI safety testing and evaluation methods.
• Joint research: The UK and Canada will work together to identify priority areas for collaborative research, including the emerging field of “Systemic AI Safety,” which focuses on safeguarding societal systems impacted by AI deployment. This research will involve collaboration with the US AISI as well.
• Professional exchanges: The partnership will enable exchanges and secondments between researchers in both countries, developing a cross-border flow of expertise.
• Shared resources: The UK AISI will share access to its AI research resources with the Canadian AISI to support joint research projects.
• International Collaboration: The UK and Canada, alongside the US AISI, aim to develop a program to accelerate research in “Systemic AI Safety.”
• Global Standards: Both countries are committed to establishing international AI safety standards and fostering open dialogue on AI risks. They will collaborate on future international reports on AI safety and regularly exchange information regarding standard development.

This collaboration aligns with the Bletchley Declaration, an international agreement urging responsible AI development. This partnership also builds on a Memorandum of Understanding (MoU) signed in January 2024, where the UK and Canada agreed to collaborate on AI across four key areas: access to computing power for AI, sustainable AI infrastructure, joint research projects on AI, and developing talent in the field.