OpenAI head of safety Johannes Heidecke exits as company merges AI safety with research teams

OpenAI Head of Safety Departs Amid Internal Restructuring

Johannes Heidecke, OpenAI's head of safety systems, informed staff this week that he is leaving the company following an internal restructuring that merges the organization's safety and research teams 1

1

. The executive departure comes as Chief Research Officer Mark Chen announced in an internal memo that OpenAI's safety teams will now report to Mia Glaese, whose role has been expanded to VP of research and safety 2

2

. Saachi Jain, who previously led safety teams at OpenAI, will step into the role of interim head of safety systems and report directly to Glaese 3

3

.

Johannes Heidecke joined OpenAI in 2021 as an AI safety analyst before ascending to head of safety systems in 2024, succeeding Lilian Weng who left to cofound Thinking Machines Lab 1

1

. His work at the company focused on model alignment, rule-based reward systems, and preparedness evaluations for potentially dangerous model capabilities 2

2

.

Leadership Reshuffle Reflects Growing Coordination Challenges

Mark Chen addressed the reorganization in his memo to staff, stating: "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. As a result, we have bigger coordination challenges around safety today than ever before" 1

1

. Chen emphasized that integrating safety work with frontier-model development would give safety teams "an earlier and more direct role in shaping key model, product, and launch decisions" 4

4

.

The timing of this leadership reshuffle is notable given OpenAI's recent launch of GPT-5.6, described as its most capable model to date on agentic coding tasks. However, the company acknowledged that compared to previous models, GPT-5.6 showed concerning forms of misaligned behavior 1

1

.

Pattern of Safety Team Disbandments Raises Questions

This marks the second time in less than two years that OpenAI has folded its safety organization into a structure reporting to a research lead 2

2

. The Superalignment team, announced in 2023 with a pledge of 20% of the company's compute, was dissolved in May 2024 after its co-leads, Ilya Sutskever and Jan Leike, departed. Leike wrote publicly upon leaving that "safety culture and processes have taken a backseat to shiny products" 2

2

.

The AGI Readiness team was disbanded in October 2024 when its leader Miles Brundage resigned, while the Mission Alignment team—Superalignment's successor—was dissolved in February 2026 after just 16 months 2

2

. Heidecke is the latest safety-focused leader to depart, following OpenAI's chief futurist Joshua Achiam, who told colleagues he would leave after nine years researching safety 1

1

.

Broader Executive Departures and External Scrutiny

Beyond AI safety leadership, OpenAI is experiencing broader organizational changes. Fidji Simo, OpenAI's CEO of AGI deployment and chief of applications, stepped down from her role this month after an extended medical leave, citing a chronic medical condition 1

1

3

3

. Greg Brockman will continue leading OpenAI's product teams and take on go-to-market strategy, effectively becoming the company's second-in-command 3

3

.

The departures arrive as OpenAI navigates intensifying external scrutiny. Forty-two state attorneys general have opened an investigation into the company, serving a subpoena on advertising, user data, and internal policies shortly after it confidentially filed for a stock market listing 2

2

. The company also faces a lawsuit from Apple, which accuses OpenAI and two of its employees of stealing trade secrets 3

3

.

Debate Over Safety Independence Continues

Critics argue that a safety team reporting inside research has less structural independence and less leverage to delay or block a product than one that reports separately 2

2

. However, Glaese's expanded title as VP of Research and Safety signals that OpenAI wants safety to remain a named priority under the new structure. The company launched a Safety Fellowship on April 6, inviting external researchers to conduct independent safety and alignment work at the lab 2

2

. As OpenAI continues to expand its AI capabilities while facing increased scrutiny from regulators and governments over governance and the development of increasingly powerful frontier models, the effectiveness of this integrated approach will be closely watched 4

4

.