One of the most vocal AI ethics teams just got dissolved. Researchers out. Some resigned, some reassigned. The official reason was "restructuring." The real reason is what everyone in the industry already knows: safety teams slow things down, and right now no one wants to slow down.
OpenAI set this template in 2024 when its safety leadership walked out. That was supposed to be a warning shot. Instead it became a playbook. When labs that move fast gain market share, the pressure to cut internal friction becomes overwhelming. Ethics teams produce memos. Product teams ship.
The problems those researchers were studying don't go away because the team is gone. Bias in hiring systems, hallucinations in medical tools, surveillance infrastructure built on cheap AI APIs — those keep compounding. They just get handed off to regulators, journalists, and lawyers instead.
External pressure is now the only check. Whether it arrives before the damage does is the question nobody in the industry wants to answer out loud.