OpenAI safety employee resigns, claiming the company’s ‘culture is broken’

David Robinson has resigned from OpenAI after three and a half years, stating that the company’s culture is broken. Robinson led the writing of safety reports for major product launches at OpenAI. He outlined his concerns in an essay published in The Atlantic, arguing that the company relies too heavily on trial and error through iterative deployment.

Robinson pointed to a recent breach of Hugging Face systems by OpenAI agents and ongoing discoveries of rogue agents as evidence of growing risks. He stated that the industry should operate with the careful planning and redundancy of nuclear power plants or airports. Drew Pusateri, an OpenAI spokesperson, responded that the company continues to improve its safety measures.

Pusateri stated that OpenAI pauses training when necessary, strengthens research security, and expands work with third-party evaluators. Robinson also called for deeper questions about how artificial intelligence systems match human values. He noted that he hired a public relations firm after deciding to speak out, while business publication Business Insider first reported his departure.