David Robinson has left OpenAI. Business Insider reports confirmation from a company spokesperson. Robinson worked on safety transparency and the system cards documenting new models' capabilities and limitations.
In an October 3 Atlantic essay, he explains his resignation: trial and error is insufficient as AI capabilities grow. He calls for stronger safeguards and alignment evidence. The essay says OpenAI considers its safety practices adequate.
The sources describe a narrower role than some headlines suggest: safety transparency is not overall leadership of all safety work. Business Insider also mentions separate staffing decisions; that does not establish a shared reason for leaving.
AI Incidents records the public warning as a risk signal. Its publication is verified; the underlying risk judgments remain Robinson's. This article creates no additional incident and does not change the incident timeline.