Vibeleaderboard
SS

ai lab

Also indexed as safe-superintelligence

Safe Superintelligence

Safe Superintelligence concentrates on developing capability and safety together, with safe superintelligence as its stated product goal.

Research dossier · Reviewed 2026-09-21

Research and release status

No public model release was verified on the reviewed mission and updates pages. The stated product goal is not evidence of achieved capability.

A single research objective

SSI describes safety and capabilities as joint technical problems and says it has organized its business around one eventual product. Its mission emphasizes freedom from near-term product cycles. The company lists offices in Palo Alto and Tel Aviv. [1]

Leadership and infrastructure

A July 2025 update names Ilya Sutskever as CEO and Daniel Levy as president after Daniel Gross's departure. A July 2026 update announces an NVIDIA partnership intended to expand compute capacity tenfold. These establish organizational and infrastructure changes, not a public demonstration of superintelligence. [2]

The evidence still needed

The next substantive assessment depends on technical publications, safety evaluations, and details of any released system. We track the company's own updates and official account without treating online release rumors as a confirmed launch. [1][2]

Sources

  1. [1] Company missionSafe Superintelligence · accessed 2026-09-21
  2. [2] Company updatesSafe Superintelligence · accessed 2026-09-21