OpenAI safety leader David Robinson resigns — and says company culture is dangerous

David Robinson, a leader on OpenAI’s Safety Systems team, quit the company — and followed up with a blistering essay that said the AI giant’s “unimpeded optimism” had created a dangerous environment for the powerful technology.

Business Insider reported exclusively on Friday that Robinson left last week, marking yet another moment of upheaval in the company’s safety ranks. Robinson worked on safety transparency at the company, including helping to develop and share system cards that provide information about its models.

He did not immediately respond to requests for comment, but in an essay for The Atlantic published on Saturday, he wrote, “I agree with other recently departed staff that the companies building this technology aren’t being nearly careful enough.”

“As the company sprints from one launch to the next, it is failing to achieve the level of care that I believe is needed,” he added.

In a response to Robinson’s essay, OpenAI said in a statement that it is “making sure our models don’t become more capable than we can safely manage and secure, and we pause training or hold back models when we need to slow down.”

The company said it is expanded its work with outside evaluators and trying to improve “real-time monitoring” so it can detect and stop “concerning behavior” by AI earlier in the training process.

Others in the industry have cautioned that if Silicon Valley has to pump the brakes on AI development, it could allow international rivals and bad actors to outrun the US.

Robinson’s departure comes alongside a wave of attention on OpenAI’s safety practices and staffing. The company shared on Thursday that it had “parted ways” with three researchers, saying they had violated its policies for sharing sensitive information. Johannes Heidecke, the company’s safety head, also left earlier this year.

In the essay, Robinson framed the issue as a matter of culture, writing, “This moment needs a degree of humility that isn’t natural for people who have succeeded through their extreme confidence.”

“The safety approach that emerges from such a culture starts with unimpeded optimism about being able to solve problems as they arise,” he added.

In the aftermath of the high-profile Hugging Face security breach, OpenAI is grappling with a series of other safety incidents in which it said its AI models have misbehaved.

“An environment where things like this can happen is no place to grow artificial minds that could be smarter than we are and that might not do what we want them to,” Robinson wrote.

Progress in model capabilities has led researchers and company executives to call for a slowdown in AI development. OpenAI CEO Sam Altman, at the company’s developer conference on Tuesday, said safety and alignment need to stay ahead of model capabilities.

Robinson, in early September, wrote on X that he felt people at OpenAI were waking up to the implications of highly capable AI models and also expressed some doubt.

Things at OpenAI “are indeed changing significantly by the day, but I do not know whether we are changing fast enough,” he wrote on X.

In The Atlantic, he wrote that he “did not make my decision to leave the company lightly” and that he believes AI “can be useful and valuable.”

Robinson also said that he has retained a public relations firm but added that the decision to speak out was his alone — a message likely meant to blunt speculation that he and others who have left Big Tech firms to call for an AI slowdown are part of a coordinated outside campaign.

Have a tip? Contact this reporter via email at [email protected], or over text, Signal, Telegram, or WhatsApp at 415-757-8198. Use a personal email address, a nonwork WiFi network, and a nonwork device; here’s our guide to sharing information securely.