Who holds the kill switch when the person overseeing it goes on medical leave?
That is the uncomfortable question now confronting OpenAI after Fidji Simo, the company's CEO of AGI deployment, announced she will step away for several weeks due to a neuroimmune condition. The timing could not be worse. OpenAI is navigating its most consequential deployment phase since the company's founding — rolling out systems that the company's own researchers describe as approaching human-level reasoning across a widening range of domains.
Simo's role was specifically designed to bridge the gap between OpenAI's safety-conscious research division and its aggressive commercialization ambitions. She arrived in mid-2024 from Instacart with a mandate to build the company's consumer-facing super app, a project now led by president Greg Brockman during her absence, according to an internal memo viewed by The Verge. On the business side, CSO Jason Kwon, CFO Sarah Friar, and CRO Denise Dresser will jointly manage operations.
The structural design was deliberate: Simo sat at the intersection of capability and accountability. Her departure raises questions that go beyond personnel logistics. Brockman, who co-founded OpenAI alongside Sam Altman, has publicly championed the importance of iterative deployment — releasing AI gradually while monitoring for risks. But critics inside and outside the company have long argued that the pressure to ship frontier models creates institutional incentives to move faster than safety reviews can keep pace.
"The people who understand the technical risks are often not the same people making the go-to-market decisions," said one researcher who left the company last year and asked not to be named. "Simo was supposed to be the bridge. Now the bridge is out."
The company's communications team disputes that framing. In a statement, OpenAI emphasized that safety oversight is distributed across multiple teams and governance bodies, not concentrated in any single executive. "Our safety processes operate independently of any individual's availability," the statement read. "We continue to deploy systems through rigorous evaluation stages regardless of personnel changes."
That explanation may satisfy investors focused on quarterly metrics. It is less likely to reassure the growing coalition of academics, policymakers, and civil society groups who have demanded greater transparency into how frontier AI companies make deployment decisions. OpenAI has committed to external red-teaming and third-party audits, but the specifics of who can halt a launch — and under what conditions — remain proprietary information.
Simo is not the only departure adding turbulence. Chief Marketing Officer Kate Rouch also announced her resignation this week to focus on health, per the same Verge report. Together, the exits mark the third major reorganization of OpenAI's executive suite in eighteen months, following the brief ouster of Altman in late 2023 and the subsequent restructuring that elevated Simo to her newly created role.
The AGI deployment office was itself a novel construct — a position that did not exist at any prior AI laboratory because no prior AI laboratory had reached this threshold. OpenAI created it to signal seriousness about managing the transition period between narrow AI and systems that could pose novel risks if deployed carelessly. That office now operates without its chief architect, for an undefined period, at the precise moment when the first systems from that office are reaching customers.
Brockman returns to a role he once held before the company restructured around its for-profit arm. Whether he can simultaneously champion safety culture while accelerating the super app timeline will define the next phase of the company's public accountability. Simo's medical leave is temporary. The precedent it sets about who controls the dial may not be.