What Safe Superintelligence does
Safe Superintelligence Inc. (SSI) is a U.S.-based AI safety research lab whose stated sole focus is “building safe superintelligence,” framed as the central technical problem SSI exists to solve. SSI positions its work as a “straight-shot” program—building toward superintelligence while keeping safety ahead of capabilities—and says it is not pursuing distraction from product cycles or short-term commercial pressures. SSI also describes itself as operating from offices in Palo Alto and Tel Aviv, recruiting technical talent for research and engineering in support of this mission.
Publicly, SSI has not launched a consumer or enterprise AI product or model; its public-facing “product roadmap” is described as the mission itself. The company’s visible operational differentiators are (1) a research-first business model centered on long-horizon safety work, and (2) partnerships for advanced AI compute capacity, including Google Cloud TPU support for its research and a later Nvidia partnership aimed at scaling compute on Nvidia’s Vera Rubin GPU platform.
As of mid-2025, SSI’s leadership shifted: co-founder Daniel Gross departed the company as of June 29, 2025, with Ilya Sutskever stepping into the CEO role and Daniel Levy becoming president. By July 2026, SSI publicly re-emerged from stealth with a long-term partnership announcement with Nvidia to scale its research.