David Robinson, a former OpenAI safety leader who was involved in writing safety reports accompanying the company’s product releases, has resigned and publicly raised concerns about the culture surrounding the development of increasingly capable artificial intelligence systems. In an essay explaining his decision, Robinson argued that AI companies need to place greater emphasis on safety and organisational culture as the technology continues to advance.
His comments come amid broader discussion within the AI industry about how companies should manage the risks associated with increasingly autonomous systems. Robinson pointed to incidents involving rogue AI agents and argued that frontier AI development may require stronger safeguards, greater reliance on safety expertise from fields such as aviation and nuclear energy, and new approaches to ensuring autonomous systems can remain under human control. Other figures in the field have also recently expressed concerns about the potential long-term consequences of increasingly powerful AI.
OpenAI has said it is continuing to strengthen its safety and security practices and is taking steps to slow or pause development when necessary. The company pointed to recent decisions to pause training and hold back a model following safety concerns during internal testing. The developments highlight an ongoing debate across the technology sector: how AI companies can continue advancing rapidly while ensuring that safety, oversight and responsible deployment remain central to the development process.
