UK AI Safety Institute (AISI)
Government body evaluating frontier models pre-deployment
The UK AI Safety Institute (renamed AI Security Institute in 2025) is the world's first state-backed frontier-model evaluation body, running pre-deployment safety testing agreements with OpenAI, Anthropic, and Google DeepMind. It exports the evaluation methodology adopted internationally and is the institutional counterpart labs grant early model access to.
First-of-kind
State frontier-model evals
Partners
OpenAI, Anthropic, DeepMind
Renamed
AI Security Institute (2025)
How it fits the stack
UK AI Safety Institute (AISI) with what it depends on (above) and what it feeds (below). The figure renders as a crawlable diagram and upgrades to an interactive 3D graph as it scrolls into view.
UK AI Safety Institute (AISI) in the AI stack. UK AI Safety Institute (AISI) with its immediate upstream dependencies (top) and downstream dependents (bottom) in the AI value chain. Hover a node in 3D, or read the full relationships below.
Graph data (text) — 5 entities, 4 relationships
- UK AI Safety Institute (AISI) —partners with→ Anthropic
- UK AI Safety Institute (AISI) —partners with→ Claude (Opus / Sonnet)
- UK AI Safety Institute (AISI) —partners with→ Google DeepMind
- UK AI Safety Institute (AISI) —partners with→ OpenAI
Context — capital, rivals, policy · · 4