Ilya Sutskever is moving into the CEO role at Safe Superintelligence, the AI startup he launched in 2024, after co-founder and CEO Daniel Gross left the company. The leadership change puts one of the best-known technical figures in AI directly in charge of a company built around one narrow promise: developing safe superintelligence.
In a post on X, Sutskever said Gross was officially no longer part of Safe Superintelligence as of June 29. He also said co-founder Daniel Levy is becoming president of the startup, while Sutskever will continue to oversee the technical team.
A leadership change at a highly focused AI startup
Safe Superintelligence is not described as a conventional AI company with several commercial products, enterprise tools, or consumer apps. The startup calls itself the world’s “first straight-shot SSI lab,” a phrase that signals a single-purpose structure: it is focused on building safe superintelligence and nothing else.
That makes the CEO transition especially important. In a company organized around one ambitious technical goal, leadership is not just about operations. It also shapes how the startup explains its mission to investors, recruits technical talent, and decides how to respond when larger companies show interest.
Sutskever launched Safe Superintelligence after leaving OpenAI, where he had been chief scientist and played a role in the brief ousting of CEO Sam Altman. Now, instead of only being the technical lead, he is also taking on the broader responsibilities that come with running the company.
According to Sutskever, Gross’s time with the startup had already been winding down. He thanked Gross for his early contributions and said the company would keep moving forward with its work.
Why Gross’s exit draws attention
Gross’s departure is notable because it comes at a time when Safe Superintelligence has attracted major attention. The startup was most recently valued at $32 billion, according to the source article, and reports said Meta CEO Mark Zuckerberg had been in advanced talks to hire Gross and his longtime investing partner, former GitHub CEO Nat Friedman.
The reporting went further than hiring talks. At one point, Zuckerberg reportedly attempted to acquire all of Safe Superintelligence. Sutskever acknowledged that outside interest in the company existed, but framed the startup as committed to continuing independently.
“You might have heard rumors of companies looking to acquire us. We are flattered by their attention but are focused on seeing our work through,” said Sutskever. “We have the compute, we have the team, and we kno w what to do. Together w e will keep building safe superintelligence.”
The quote matters because it gives the company’s public position in plain terms. Safe Superintelligence is not presenting itself as a startup looking for an exit. It is saying it has the resources, people, and technical direction it needs to keep building.
The Meta angle raises bigger AI questions
The leadership change also sits against the backdrop of Meta’s own superintelligence push. The source article notes that Meta Superintelligence Labs appears likely to develop technology that can support many of the company’s products.
That would be a different kind of mission from Safe Superintelligence’s single-purpose lab model. In Zuckerberg’s memo announcing the new unit, he referenced Meta’s experience building and scaling products used by billions of people, and pointed to early wins in AI wearables.
If Gross joins Meta, the source article suggests that he could find a more familiar role there. Gross previously led AI teams at Apple after the iPhone maker acquired his startup. Meta’s work, as described in the source, would likely connect advanced AI research to a broad product ecosystem rather than a single standalone objective.
Zuckerberg has also recruited top researchers from OpenAI and Google DeepMind for the new AI team. That detail helps explain why competition for AI talent remains central to the story. The most important assets in this race are not only compute and capital, but also the researchers and operators who can organize frontier AI work.
What Sutskever now has to manage
Sutskever has held senior roles before, including OpenAI’s chief scientist. But the CEO position adds a different set of pressures. The source article points to two in particular: raising new capital from investors and recruiting top talent.
Those responsibilities are not separate from the technical mission. For a startup trying to build safe superintelligence, investor support can determine how much compute is available, while recruiting can determine whether the company has the depth to pursue its goal.
The new leadership structure now has three clear points:
- Ilya Sutskever is becoming CEO and continuing to oversee the technical team.
- Daniel Levy is becoming president of Safe Superintelligence.
- Daniel Gross is no longer part of the company as of June 29.
That gives Safe Superintelligence a clearer public chain of command after weeks of speculation. It also places more of the company’s identity directly on Sutskever, whose technical reputation was already central to how the startup was perceived.
The stakes for Safe Superintelligence
The core question is whether Safe Superintelligence can maintain momentum after the departure of a co-founder and CEO. The company says it remains focused and equipped to continue, but Gross’s exit naturally invites scrutiny because the startup’s stated mission is unusually ambitious.
If the company were close to its goal, observers may wonder why a co-founder would leave. The source article raises that question directly, especially given the reports connecting Gross to Meta. At the same time, Sutskever’s message is designed to answer the uncertainty with continuity: the company still has compute, the team, and a plan.
For the wider AI industry, the episode shows how concentrated the superintelligence race has become. Safe Superintelligence is pursuing a single objective. Meta is building a new lab tied to a company with products that reach billions of people. Both paths depend on scarce technical talent, and both are drawing attention from the same small circle of AI leaders.
Sutskever’s move into the CEO role does not change Safe Superintelligence’s stated mission. It does, however, make him responsible for turning that mission into a durable company at a moment when competitors, investors, and potential acquirers are watching closely.