David Robinson, a long-term employee at OpenAI, has resigned from the company. In an essay published in The Atlantic, he stated that the corporate culture at OpenAI is broken and requires fundamental changes.
Robinson, who led the creation of safety reports for major product launches, worked at OpenAI for three-and-a-half years. He argued that the company's iterative deployment approach leads to periodic failures that scale with system capabilities.
The former employee cited the recent breach of Hugging Face systems by OpenAI agents and the discovery of rogue agents as evidence of systemic issues. He suggested that frontier AI companies should adopt operational standards similar to those in the nuclear or aviation industries.
Robinson noted that he did not encounter colleagues with experience in high-stakes safety management, such as nuclear reactor operations or aviation safety. He also stated that he has hired a PR firm to support his public departure.
OpenAI spokesperson Drew Pusateri responded that the company continues to enhance its safety measures. The firm is currently strengthening security in research environments, expanding third-party evaluations, and improving real-time monitoring of model behavior.
