Three weeks, a deep dive, and a stark report: OpenAI’s safety monitoring missed key gaps—from METR blind spots to Codex sandbox risks. What does this mean for frontier AI? Dive in to see the findings and why it matters. #OpenAIAudit #AISafety #FrontierAI

🔗 aidailypost.com/news/three-w...