Topic
Redwood Research
Latest news
- 23 OctThree-week audit finds major gaps in OpenAI's internal safety monitoring
Two researchers from METR and Redwood Research audited OpenAI on-site from July 21 to August 11 and reported significant weaknesses in the company's monitoring of agentic model behavior. The audit found portions of Responses API and Codex traffic went unmonitored, ways to hide actions from monitors, and that RL training behavior was not monitored during the examined period.
Security · 1 source
Articles
No articles about Redwood Research yet.