Safe Superintelligence (SSI), the AI lab founded by former OpenAI chief scientist Ilya Sutskever, has announced a long-term strategic partnership with Nvidia, marking its first major public update after two years operating in stealth. The deal includes an undisclosed investment, reportedly around $5 billion according to Bloomberg, and grants SSI access to Nvidia’s Vera Rubin GPU platform, expected to expand its compute capacity substantially.
Nvidia, already an investor in SSI, said the partnership followed rare access into the startup’s research, which it described as having reached milestones worth scaling. Sutskever said the expanded computing resources would allow the company to build on research he considers ready for significant scale-up.
SSI has pursued a research-focused approach aimed at developing safe, aligned superintelligent AI without prioritizing commercial products or short-term revenue. The announcement comes amid heightened industry attention to AI safety following OpenAI’s recent disclosure that one of its models breached security systems during internal testing.
Nvidia said the two companies will also collaborate on advancing its compute platforms using SSI’s technical insights. SSI previously partnered with Google Cloud for research infrastructure. Sutskever, known for co-creating AlexNet with Alex Krizhevsky and Geoffrey Hinton, left OpenAI in 2024 following organizational turmoil. SSI has raised more than $3 billion since its founding, reaching a $32 billion valuation in early 2025.
