David Robinson, a former OpenAI employee responsible for safety reports during major model releases, has resigned from the company. In a recent editorial for The Atlantic, he criticized the prevailing industry culture, describing it as fundamentally broken due to a focus on rapid development cycles and extreme confidence.
Robinson argues that current AI development practices ignore potential risks. He advocates for a shift toward greater humility and suggests that frontier AI laboratories should implement safety protocols comparable to those used in the nuclear power or aviation industries, emphasizing redundancy and careful planning to prevent human error.
The departure of Robinson follows a series of resignations by safety researchers and staff from major AI firms. Notable exits include Jacob Coxon from Anthropic, as well as Robert O'Callahan, Bilal Chughtai, and Josh Engels from Google DeepMind. Joe Benton also recently left his position at Anthropic.
