David Robinson, who helped develop OpenAI’s preparedness framework and oversaw safety reports for 12 frontier-model launches, publicly criticized the company’s approach to AI safety in an essay published Saturday by The Atlantic.

In the piece, titled “I Quit OpenAI Because Its Culture Is Broken,” Robinson argued that AI companies are not being sufficiently cautious as they develop increasingly advanced systems. He said companies should place greater emphasis on safety expertise and research before deploying more capable models.

Robinson specifically criticized OpenAI’s reliance on “iterative deployment,” an approach in which systems are released, and safeguards are strengthened as problems emerge. He argued that the industry has reached a point where trial and error is no longer an adequate approach to managing increasingly powerful AI.

He compared the level of caution he believes is needed to safety practices used in industries such as nuclear power and aviation, where safeguards are designed to prevent serious failures before they occur.

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

His criticism comes amid broader debate over the pace of AI development, including scrutiny of safety failures involving OpenAI and rival AI company Anthropic. Robinson also warned that AI capabilities are advancing faster than researchers’ understanding of alignment, the field focused on ensuring AI systems behave in accordance with human goals and values.

OpenAI defended its approach, saying the company monitors the safety of its models and is willing to slow development when necessary.

“We’re 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,” an OpenAI spokesperson said.