I Quit OpenAI Because Its Culture Is Broken
David Robinson’s first-person account of his resignation puts safety staffing and organizational pace at the center of the debate.
TL;DR
- David Robinson said he resigned from OpenAI after three and a half years, during which he led safety-report writing for product launches.
- In his account, Robinson argued that the company’s fast deployment cycle and reliance on iterative fixes leave gaps in safety practice.
- OpenAI told The Guardian that it continues to strengthen safety and security practices and pauses or holds back models when needed.
In a first-person essay, David Robinson said he resigned from OpenAI and had led the writing of safety reports accompanying major launches. He argued that the company’s pace and “trial and error” approach can allow failures to recur as systems become more capable. [1] [2]
The Guardian separately reported Robinson’s departure and included OpenAI’s response that the company is strengthening safety practices and can pause training or hold back models. Robinson’s assessment of the company’s culture is his own account, not an independently established finding. [1] [2]
Why it matters
The resignation adds a first-hand account to wider debate about how frontier AI labs organize safety work as they deploy more capable systems. The concrete disagreement is about organizational practice and safeguards; the cited material does not establish a general probability of future harm.
Editor's note
The Atlantic essay is Robinson’s first-person account; The Guardian provides separate reporting and OpenAI’s response. Claims about culture and risk are attributed to Robinson.