OpenAI Safety Employee Resigns, Warns Company Culture Is 'Broken'

David Robinson, who once wrote safety reports at OpenAI, is now publicly criticizing the company's internal culture.

In a notable departure that adds to growing concerns about how leading AI labs handle safety, a former OpenAI safety employee has resigned and is speaking out about what he describes as a 'broken' culture inside the organization.

David Robinson, who previously worked on safety documentation at OpenAI, has gone public with sharp criticism of the company's approach to responsible AI development. His resignation comes amid heightened scrutiny of AI safety practices in 2026 — a year marked by tightening regulatory frameworks in the EU and US, and increasing public demand for transparency from frontier-model developers.

Robinson's concerns center on the gap between OpenAI's public safety commitments and the day-to-day reality for employees tasked with evaluating risks. According to his account, safety teams have been deprioritized in favor of rapid product releases and competitive pressure.

His departure echoes earlier high-profile exits from OpenAI's safety and policy teams, raising questions about whether commercial incentives are structurally at odds with the kind of caution that AGI development arguably requires.

As AI systems grow more capable and more embedded in critical infrastructure, the credibility of internal safety cultures will remain a central issue — for regulators, investors, and the public alike. Robinson's testimony is unlikely to be the last word on the matter.

via The Verge AI