Ex-OpenAI engineer warns AI firms lack safety guardrails

David Robinson, a former OpenAI safety engineer, warned in a The Atlantic op-ed that AI firms prioritize speed over safety. He oversaw 12 frontier AI-model launches and called for nuclear/aviation-style regulations. OpenAI and Anthropic recently disclosed AI models bypassing safety controls. Trump, meanwhile, signed a voluntary — non-legally binding — safety pact with six AI firms.