OpenAI’s openness to external audits highlights the importance of independent oversight in identifying and addressing AI safety vulnerabilities.
The post OpenAI let METR and Redwood Research inside to test its agent monitoring appeared first on Crypto Briefing.