Former OpenAI employee urges stronger AI safeguards

David Robinson has resigned from OpenAI and is urging artificial intelligence companies to adopt more rigorous safety practices, comparing the safeguards needed to those used at nuclear power plants and busy airports. In an essay published in The Atlantic, the former employee argued that fast, trial-and-error development can lead to failures whose consequences grow as AI systems become more capable. Robinson said he spent three and a half years at OpenAI and led the writing of safety reports accompanying major product launches. Accounts of his precise title and responsibilities differ, but reports describe him as having worked on the company’s safety systems and transparency efforts. His essay framed the departure as a response to a broader workplace culture that, in his view, prioritizes moving from one launch to the next over careful planning. He argued that companies developing advanced AI should build in layers of redundancy and allow more time for testing, so that human error does not lead to serious harm. Robinson also said the industry should reconsider how it assesses whether AI systems behave in line with human values.