UK, US and Canada to collaborate on AI and cyber security

The UK government has signed a collaboration agreement with the US and Canada, which will see the countries work together to leverage new artificial intelligence (AI) and cyber security technologies.

The Ministry of Defence’s (MoD) Defence and Science Technology Laboratory (DSTL) will lead the work in the UK, while the US Defence Advanced Research Projects Agency (DARPA) and Defence and Research Development Canada (DRDC) will manage the work in their respective countries.

The countries will jointly research, develop, test and evaluate new technologies for AI, cyber and information domain-related technologies.

MoD Science and technology director Nick Joad said that international research collaborations with US and Canada are “some of our most vital and enduring partnerships”, adding: “This agreement cements our collective commitments to advancing emerging cyber security technologies such as cyber security and AI to enhance the defence and security of our nations.”

The countries hope to be able to exploit new methodologies, algorithms and tools by developing real-world concepts of operations. The collaboration will focus on current and future challenges in a changing geopolitical landscape, and aims to bolster the defence and security of the three countries.

DARPA director Stefanie Tompkins said that the three nations are “stronger together than separately”, adding: “The trilateral collaboration is a big step toward enhancing our understanding in the outlined research and development thrust areas. Working with our international partners on science and technology helps us all leverage each other’s individual strengths to develop much greater collective capability.”

The collaboration agreement also covers plans to reduce technological risks so new capabilities can be introduced as quickly as possible.

There is already a project underway working to train AI to autonomously defend networks against persistent cyber threats. Other areas the collaboration will look into include human-AI teaming, defining and creating trustworthy AI systems, protecting and detecting attacks on the information domain, and producing new tools and techniques – such as rapid certification of software – to increase the resilience and security of systems.

Canada’s DRDC assistant deputy minister Jaspinder Komal said that due to the rapid evolution of AI in “dual-use technologies with civilian and military applications, we’ve identified this as a priority for defence research”.

“We’re pleased to be working with our allies in the UK and US through this trilateral agreement, which will potentially enable a multitude of cooperative research projects in the cyber and information domains,” said Komal.

In February 2024, the UK signed two other collaboration agreements with Canada, joining efforts to look at AI research and development (R&D) – in particular, AI compute, which is an essential component in the development of AI, as well as a focus on science and innovation.

Earlier this month, the UK also signed the world’s first “legally binding” treaty on AI and human rights, which commits to implementing safeguards against various threats posed by the technology.