Agent Safety Blame Game Grows as OpenAI Faulted for Silent Security Posture

A purported OpenAI security insider's remark that agent risk 'isn't just the sandbox' sparked debate, and cryptographer Matthew Green criticized OpenAI for having no one publicly responsible for agent security communication amid rising breakout incidents.

2026-09-29 ~ 2026-09-30 · 3 related posts

Full story(8 episodes)→

1 near-duplicate retellings: matthew_d_green