Safe Superintelligence Inc. (SSI) is an AI research company founded in June 2024 by Ilya Sutskever, the former chief scientist and co-founder of OpenAI, alongside Daniel Gross and Daniel Levy. The company describes itself as pursuing a single objective, building a safe Superintelligence, with no other products, and states it will not release intermediate commercial systems before reaching that goal.
Founding
Sutskever left OpenAI in May 2024, several months after playing a central role in the November 2023 OpenAI board crisis that briefly removed Sam Altman as CEO. He had previously led OpenAI's Superalignment team focused on aligning future superhuman AI systems, a team that was effectively dissolved after his and co-lead Jan Leike's departures. SSI was announced with offices in Palo Alto, California, and Tel Aviv, Israel, reflecting Sutskever's and Gross's respective backgrounds, and the company said it had already begun assembling a small, deliberately compact technical team rather than scaling headcount broadly.
Structure and funding
SSI adopted an unusual public posture for a company operating at the frontier of AI research: it declared upfront that it would remain insulated from short-term commercial and product pressures, avoiding the cycle of product launches, benchmark competition, and quarterly revenue targets that shape labs like OpenAI, Anthropic, and Google DeepMind. Despite eschewing near-term products, SSI raised substantial venture funding, including a 2024 round reported around $1 billion and further funding through 2025 that pushed its valuation into the tens of billions of dollars, remarkable for a company with no public product, benchmark result, or revenue. Daniel Gross, a former Y Combinator partner and AI investor, departed the company's operating role in 2025 to join Meta's superintelligence effort, while Sutskever continued as CEO.
Reception
SSI's premise, that safety and capability should be pursued together from a single research agenda rather than bolted on after a capable model already exists, drew both interest and skepticism. Supporters framed it as a natural extension of ideas explored in AI alignment research and a bet that a smaller, focused team could make progress other labs' commercial obligations crowd out. Skeptics questioned whether any lab can credibly claim to have solved existential risk before building the very system that risk concerns, and noted that SSI's record valuation, absent any disclosed product or benchmark, reflected as much investor confidence in Sutskever personally as in the company's stated mission. As of 2025, SSI had not released a public model, paper, or benchmark result.