David Robinson, who spent three and a half years at OpenAI and helped write safety reports for major AI launches, has quit the company. In an essay in The Atlantic, he said OpenAI’s culture had become “broken.” His main concern was that the company was moving so fast that it did not have enough time to think carefully about safety.
Robinson said AI companies should work more like airports or nuclear plants, where multiple safety checks prevent small human mistakes from becoming major disasters. He argued that OpenAI was stuck in a nonstop sprint, releasing and improving products quickly. According to him, this left little room to pause, review risks, and make deeper changes.
His exit came soon after OpenAI reportedly dismissed three safety researchers for allegedly sharing sensitive information with an outside safety group. This has renewed concerns about whether employees can freely raise safety warnings. Earlier, former OpenAI leader Jan Leike had also said that safety was being pushed behind new products.
The larger issue is that OpenAI began as a non-profit research organisation, but has now become a powerful commercial company. Its decisions no longer affect only its business; they can shape the future of AI and society. When the organisations building the most powerful technology cannot slow down for safety, the risk is not just corporate—it is human.
WHEN AI MOVES FASTER THAN SAFETY, EVERYONE PAYS THE PRICE.
Sanjay Sahay
Have a nice evening.