UK AI Security Institute (formerly AI Safety Institute)
Government body that tests frontier models pre-deployment and researches AI security risks.
The UK's state AI-testing lab, the first of its kind, examines frontier models before launch. In 2025 it rebranded to 'security' and narrowed its remit away from bias and societal harms.
Focus
- Safety evaluations
- Transparency
- International governance
Funding, as far as we know
UK government (Department for Science, Innovation and Technology; initial £100M) - https://www.gov.uk/government/organisations/ai-security-institute
What they've actually done
Jointly evaluated OpenAI's o1 with the US AISI before release, publishing cyber/bio/software results
2024-12
Renamed to AI Security Institute, dropping bias and free-speech work to focus on security risks
14 February 2025
Hosts the secretariat of the International AI Safety Report (2026 edition)
3 February 2026
Joint UK-US preliminary cyber assessment of Moonshot's Kimi K3
2026-07