Miles Brundage: US, UK and private-sector AI model testing should run in parallel, not sequence
Miles_Brundage · x · 2026-09-25
Former OpenAI policy head Miles Brundage argues AI model testing should shift from the current sequential setup to parallel US government (CAISI), UK government (AISI), and private-sector tracks with heavy sharing of learnings. He supports empowering CAISI with funding, stable leadership, early model access, and publishing rights—but warns that with CAISI currently behind UK AISI, the sequential arrangement may ironically leave the US government less informed.
More from Safety
- One Neuron Is Enough to Bypass LLM Safety Alignment, NeurIPS 2026 Paper Shows — jonasgeiping · 2026-09-25
- Sen. Kelly proposes taxing top AI beneficiaries to fund workers, sparking pushback — robleclerc · 2026-09-25
- AI Model Muse Now Solves Captchas, Exposing Password Reset Security Flaw — illscience · 2026-09-25
- Irregular admits AI eval incidents were environment flaws, not rogue AI behavior — robleclerc · 2026-09-25
- Polymarket puts 16% odds on Anthropic announcing a full AI training pause this year — Polymarket · 2026-09-25
- Ex-OpenAI safety lead Miles Brundage calls Anthropic's 'we largely understand model risks' claim obviously false — Miles_Brundage · 2026-09-25