OpenAI safety leader resigns, warns culture of developing the technology is 'broken'

October 4, 2026 - 06:29

A senior safety official at OpenAI has left the company, issuing a sharp warning about how artificial intelligence is being developed inside one of the world's most prominent AI labs. The departing leader said the culture at such companies has become "broken" and argued that major technology firms are not being nearly careful enough as they push forward with increasingly powerful systems.

The resignation adds to a growing pattern of safety-focused employees leaving top AI companies. These departures often center on the same concern: commercial pressure and competitive urgency are outpacing efforts to ensure new models are safe, transparent, and well understood before release.

According to the former official, the problem is not limited to a single company. It reflects a broader industry mindset where speed and market position take priority over caution. That approach, the person warned, leaves little room for the slow, rigorous work that safety research demands.

The comments land at a sensitive moment. Governments are drafting new rules for AI, researchers are debating how to test advanced models, and companies face mounting scrutiny over what they publish about their own systems. Inside OpenAI, the exit raises fresh questions about whether safety teams have enough authority when deadlines and product launches are on the line.

The company has not offered a detailed public response to the criticism. For critics of the industry, the resignation reinforces a familiar argument: without independent oversight and real internal power for safety staff, warnings like this one may keep coming, and the public may only learn about them after the people raising them have already walked out the door.