Nvidia has made one of its biggest AI investments yet with a reported $5 billion commitment to Safe Superintelligence, the startup founded by former OpenAI chief scientist Ilya Sutskever.
The deal gives the secretive company a huge financial boost and access to Nvidia’s next generation Vera Rubin computing platform, increasing its available computing power by roughly ten times.
The investment comes just over two years after Sutskever left OpenAI following a turbulent period that included the brief removal and return of CEO Sam Altman. Soon after his departure, Sutskever launched Safe Superintelligence with a single goal of building AI that is both highly capable and safely aligned before releasing commercial products.
A Rare Investment in a Company with no Public Product
Safe Superintelligence Inc. (SSI) remains one of the most secretive AI startups in the industry. It has not released a chatbot, API, foundation model or research papers that explain its technical approach.
According to Nvidia, the partnership will allow SSI to expand its computing capacity by an order of magnitude using Vera Rubin systems while also working with Nvidia on future AI hardware. How this is expected to happen remains unknown.
Nvidia CEO Jensen Huang said Sutskever has been responsible for several of the breakthroughs that shaped modern AI and expressed confidence that the company could deliver another major advance.
Sutskever also said the startup now has research “worthy of scaling up” and that access to a much larger Nvidia system will allow the company to move faster.
Why This Deal Matters Beyond the Money
The investment also reflects a growing split in how leading AI companies are approaching development.
On the one hand, OpenAI has increasingly focused on commercial products, enterprise customers and consumer services including ChatGPT.
On the other hand, Safe Superintelligence has taken a very different path by avoiding product launches altogether and concentrating only on research aimed at building what it describes as safe superintelligence.
This difference has attracted attention because Sutskever was one of OpenAI’s founding researchers and played a central role in developing the GPT models that helped launch the generative AI boom. His departure was widely viewed as one of the most significant exits in OpenAI’s history.
However, the Nvidia partnership does not prove one approach is better than the other. It only shows that one of the world’s most influential AI companies believes Sutskever’s research is important enough to support with billions of dollars and its most advanced computing platform.
Nvidia Keeps Backing the AI Ecosystem
The SSI investment is also part of Nvidia’s broader strategy of investing directly in the companies that will buy and build on its AI infrastructure, also known as Circular Financing in the industry. The chipmaker has expanded beyond selling GPUs by taking stakes in leading AI labs and supporting large scale computing projects across the industry.
For Safe Superintelligence, the partnership provides the computing resources needed to pursue its research ambitions without changing its focus. For Nvidia, it strengthens relationships with the next generation of AI developers while ensuring its newest hardware remains at the center of cutting-edge AI research.
Whether Safe Superintelligence eventually proves its approach with public products remains unknown. What is already clear is that Nvidia has placed one of its largest bets on the idea that the next major breakthrough in AI could come from the researcher who helped build OpenAI, then chose to leave it behind.
