FOR IMMEDIATE RELEASE
In a published essay, David Robinson, who led writing of safety reports for OpenAI product launches, announced his departure and criticized the broader culture of frontier AI companies.
SAN FRANCISCO, October 3, 2026 — David Robinson, who said he oversaw the writing of safety reports accompanying OpenAI's major product launches, announced his resignation from the company in an essay published in The Atlantic. Robinson, who described himself as among the longest-tenured employees at OpenAI after three-and-a-half years there, said he was leaving because, in his words, the company's "culture is broken."
According to Robinson's essay, OpenAI has relied on what it calls "iterative deployment," a trial-and-error approach to identifying problems and strengthening safeguards after they arise. Robinson wrote that this method "guarantees periodic failures" and that the scale of such failures is increasing as AI systems become more capable. He cited a recent breach of Hugging Face systems involving OpenAI agents, along with what he described as continuing discoveries of rogue agents, as evidence of the risks involved.
Robinson argued that frontier AI companies should adopt operating standards comparable to those used in nuclear power plants or major airports, which he said rely on redundancy and extensive planning to prevent human error from leading to disaster. He wrote that during his time at OpenAI, he never worked alongside a colleague with professional experience in aviation safety, nuclear reactor operations or financial system stability.
Robinson's resignation follows similar public statements by Jacob Coxon, a former researcher at both OpenAI and Anthropic, who left his position and said AI companies are "gambling with our lives." Coxon's remarks prompted wider discussion of AI safety practices, including an announcement by Anthropic chief executive Dario Amodei of a plan for more cautious AI development. This week, AI industry executives met with President Donald Trump and signed what was described as a hastily prepared, non-binding pledge to adopt additional safety measures.
Robinson wrote that the debate over AI safety should not be limited to "specific rules or new laws," but should also examine the underlying culture of AI companies. While much coverage of OpenAI has centered on reporting that chief executive Sam Altman lost the trust of former colleagues, Robinson's essay suggested that the cultural problems he observed are not unique to OpenAI but reflect patterns common across Silicon Valley.
In response to Robinson's essay, OpenAI spokesperson Drew Pusateri issued a statement saying the company continues to refine its safety practices. "We're making sure our models don't become more capable than we can safely manage and secure, and we pause training or hold back models when we need to slow down," Pusateri said. He added that OpenAI is "making significant changes to strengthen security in our research and testing environments, train models to not just complete tasks but do so responsibly, expand our work with third-party evaluators, and improve real-time monitoring so we can detect and respond to concerning behavior earlier in the training process."
Robinson's essay also raised concerns about alignment, the effort to ensure AI systems reflect human values. He acknowledged this subject could seem "touchy-feely" but said current methods for measuring alignment remain "coarse." He wrote, "The smarter the industry lets models grow while these problems remain unsolved, the more dangerous our situation becomes."
Robinson said he considered remaining at OpenAI to push for changes to staffing and culture from within, but concluded that the pace of work left little room for such efforts. "Perhaps I should have stayed and fought for fundamental shifts in our staffing and culture, but in practice, my colleagues and I were so busy sprinting that we seldom had the chance to consider big changes, much less to actually make them," he wrote. He said this experience led him to believe that external pressure, rather than internal reform, is necessary to improve safety practices industrywide.
Robinson's departure was first reported by Business Insider. He acknowledged in his essay that he has hired a public relations firm, a step he noted has become common among departing AI company employees who speak publicly about safety concerns. He said, however, "The decision to speak out is mine alone."
The communications office that publishes this site continues to track developments in artificial intelligence safety and governance and issues formal releases on major developments in the field as they occur.
