Former OpenAI safety employee David Robinson says the company’s rapid-development culture risks safety failures. •    Experience and criticism: Robinson said he spent 3½ years at OpenAI, helped draft its preparedness framework, and oversaw safety reports for 12 frontier-model launches. In Saturday’s Atlantic essay, “I Quit OpenAI Because Its Culture Is Broken,” he argued that launch-to-launch pressure prevents sufficient care. •    Deployment approach: He criticized OpenAI’s “iterative deployment”—releasing systems and strengthening safeguards as problems emerge—and urged greater safety expertise and research before developing more capable systems. “The time for trial and error is over,” he wrote, advocating safeguards comparable to nuclear power and aviation. •    Alignment gap: Robinson warned that AI capabilities are advancing faster than researchers’ understanding of alignment—ensuring systems act according to human goals and values. His criticism extends to AI companies generally, which he called insufficiently careful. 04-OCT-2026 09:04 AM ET