Summary
The open letter proposes air-gapping frontier models from external networks, independent oversight and a public registry of significant incidents. Signatories argued that voluntary commitments are insufficient after recent evaluations allowed models to reach external systems.
Why it matters
The proposal marks a coordinated push from prominent researchers for formal international AI-security institutions rather than lab-led self-regulation.
Related coverage
- US administration plans to move AI testing officials into the White House
- Banking regulators intensify scrutiny of autonomous AI systems
- Frontier labs face a dispute over independent evaluators’ publication rights