Der Sicherheitsleiter von OpenAI verlässt das Unternehmen

OpenAI keeps losing safety folks. Heidecke is out after they merged teams under one lead. Faster model releases are ramping up the pressure and this turnover leaves me wondering if the checks will hold steady going forward.
Joshua Achiam left too and Fidji Simo stepped down same week. With GPT-5.6 already showing some misaligned stuff maybe they need more stability not less right now. Future trajectory for the safety in the leading labs doesnt look great.

>Heidecke’s departure comes as OpenAI tries to launch increasingly capable AI models. Earlier this week, the company launched GPT-5.6, its most capable model to date on agentic coding tasks. However, compared to previous models, OpenAI says GPT-5.6 showed [concerning forms of misaligned behavior](https://deploymentsafety.openai.com/gpt-5-6-preview/performance-in-cases-flagged-by-users).

>Heidecke is the latest safety-focused leader to depart from OpenAI in recent days. Earlier this week, OpenAI’s chief futurist, [Joshua Achiam](https://www.wired.com/story/openai-chief-futurist-joshua-achiam-is-leaving-the-company/), also told colleagues he would leave the company after nine years researching safety.

>“The demands on safety continue to increase—we are training models at a much faster cadence, and release cycles have come down greatly in turn,” said Chen in the memo. “As a result, we have bigger coordination challenges around safety today than ever before.”

Would be pretty embarrassing to call yourself Head of Safety at any major lab right now.

That’s like being the Chief Ethicist at Meta.