Artifical Intelligence Agencies: The Use and Regulation of AI in UK Intelligence Agencies
A look into how intelligence agencies in the UK might be using AI, what harms it poses and how it's currently regulated.
Key findings
- Intelligence agencies are likely using AI in their workflows for a variety of tasks including summarisation, translation, identifying key information in a "sea of data", sentiment analysis, profiling and more.
- Intelligence agencies can be exempted from a number of existing regulatory safeguards including those under data protection law, human rights and equality laws.
- Sector specific law, specifically the Investigatory Powers Act, does not live up to the demands of the new AI era and leaves serious gaps which can undermine rights.
Report
Post date
3rd November 2026
This report considers two questions: first, how are intelligence agencies using AI? And second, how is that use regulated? We focused on the UK in answering these questions.
To understand the often opaque operations of intelligence agencies, we had to rely on limited public information, and assumptions based on other countries use-cases and AI tool capability. Together, these provide us with a limited glimpse of what they are likely using it for. The report discusses these uses and unpacks some of the harms they pose.
To understand how such use is regulated we examined data protection law, human rights and equality laws, and intelligence laws (specifically the Investigatory Powers Act). The analysis highlights the gaps in regulation and notes recommendations to strengthen regulation.
We conclude with recommendations for how UK regulation of AI use by intelligence agencies could be improved, both by increasing the effectiveness of certain safeguards and drawing clear lines where AI should be prohibited, especially when it comes to automated decision making.
Our campaign
