Safety researcher: Keep open weights but pressure inference providers on KYC and monitoring
joshua_saxe · x · 2026-09-13
joshuasaxe says he long believed openness and diffusion was the right path to a good cyber equilibrium in the AI era, but is now reconsidering. His loosely held proposal: keep supporting open weights, but put heavy pressure on inference providers (Together, Fireworks, and closed labs alike) to enforce know-your-customer checks and monitoring, filtering out bad actors who lack the capital to self-host frontier models. Quoted context adds a defense-in-depth take: combine redteaming to find bugs with deployed sentry agents monitoring live threats, including social engineering.
Related event: Cybersecurity Experts Debate AI Offense-Defense and Inference-Layer Rules(4 posts)→
More from Safety
- AI memo: China to restrict exports of 23 AI technologies, plus Kimi K3.1 Pro and Genie 4 — dralandthompson · 2026-10-31
- Anthropic adds invisible text watermark to Claude outputs starting Nov 3 — petesena · 2026-10-31
- LessWrong essay: AI lab employees are 'microdosing' their org's spirit, argues researcher — jankulveit · 2026-10-31
- Stanford prof publishes lab AI policy: no AI-authored text in papers, ever — ChrisGPotts · 2026-10-31
- Security researcher: air-gapped AI agents are only possible with open-weight models — wunderwuzzi23 · 2026-10-31
- Podcast: Is Mistral Large 4 the West's best open model? Plus AI cyber threats — The Cognitive Revolution · 2026-10-31