Ilya Sutskever's SSI reportedly in talks to raise at $30B valuation — Sequoia and a16z both involved
Mar 5, 2025
Key Points
- Safe Superintelligence, Ilya Sutskever's AI safety startup, is raising at a $30 billion valuation with Green Oaks leading and Sequoia and Andreessen Horowitz co-investing, a sixfold jump in under a year.
- Sequoia and a16z are simultaneously backing xAI, Elon Musk's competing superintelligence play with opposite strategy, suggesting venture capital treats superintelligence bets as a portfolio play rather than mutually exclusive.
- SSI's valuation rests entirely on Sutskever's technical pedigree and safety philosophy with zero announced products or revenue, signaling venture capital now prices founder credibility alone as sufficient justification in the superintelligence race.
Summary
Safe Superintelligence, Ilya Sutskever's AI safety-focused startup, is in talks to raise funding at a $30 billion valuation. That represents a sixfold jump from its prior $5 billion valuation. Green Oaks Capital is leading the round, with Sequoia and Andreessen Horowitz also investing. The company appears to be moving from zero to $30 billion in less than a year.
Sutskever left OpenAI as chief scientist to found SSI with a stated mission to build superintelligence before releasing any products. The company operates with extreme operational security. Job candidates leave phones in Faraday cages before interviews. The website contains only a 223-word mission statement. Staff are drawn from promising technologists Sutskever can mentor rather than known Silicon Valley names. Offices sit in Silicon Valley and Tel Aviv.
Both Sequoia and a16z are invested in Elon Musk's xAI, which pursues a nearly identical mission—building frontier foundation models—but with opposite strategy. xAI prioritizes rapid product release and consumer adoption, embedding Grok in X and pursuing Tesla integration. SSI commits to stealth development until superintelligence is ready. The two firms backing both companies suggest the venture capital market treats competing superintelligence plays as a portfolio bet rather than mutually exclusive bets.
Sutskever's track record at OpenAI carries significant weight. He recognized the Transformer architecture's importance, built early GPT architectures, and later developed test-time inference and reinforcement learning approaches. His ability to recognize algorithmic breakthroughs gives credibility to the SSI bet. Given OpenAI's trajectory toward agents and the broader question of what comes after the LLM era, he may be working on architectures or paradigms beyond large language models.
The valuation rests entirely on founder reputation and technical pedigree, not traction. SSI has no announced products, revenue, or deployment. The round shows how venture capital now prices raw technical credibility and safety-aligned philosophy as sufficient justification for multi-billion valuations in the superintelligence race. OpenAI, by contrast, is a more mature AI company with $4 billion in annual recurring revenue and projections to reach $11 billion by year-end.