Safe Superintelligence Inc. (SSI)

What SSI is

Safe Superintelligence Inc., usually shortened to SSI, is an American artificial intelligence company with offices in Palo Alto and Tel Aviv.^[2] It states that it has one goal and one product: a safe superintelligence.^[2] The company says SSI is its mission, its name and its entire product roadmap, because it is its sole focus.^[2] In ordinary usage, superintelligence refers to machine intelligence that vastly exceeds human ability across a wide range of cognitive tasks; SSI places that target, rather than a commercial product line, at the centre of the company.

SSI describes itself as the world's first "straight-shot SSI lab" — a phrase meant to convey that it works toward superintelligence directly instead of shipping intermediate products along the way. It says that building safe superintelligence "is the most important technical problem of our time", and that it will advance capabilities while keeping safety ahead of them.^[2] Consistent with that framing, it has not announced any product or model.^[1]

Origins

SSI was founded on 19 June 2024, when Ilya Sutskever announced it on X with co-founders Daniel Gross and Daniel Levy, about a month after he left OpenAI.^[1]

Offices

The company's Palo Alto office is at 317 University Avenue, and it leased offices in Tel Aviv in February 2025.^[1] It describes itself as an American company with offices in those two cities.^[2]

Funding

In September 2024, SSI raised $1 billion from SV Angel, DST Global, Sequoia Capital and Andreessen Horowitz at a valuation of $5 billion.^[1] A March 2025 round led by Greenoaks Capital valued the company at $30 billion — at a time when it had no revenue and about 20 employees.^[1]

Compute and partnerships

In April 2025, Google Cloud agreed to provide TPUs for SSI's research.^[1] On 27 July 2026, SSI and Nvidia announced a strategic partnership intended to expand SSI's compute by an order of magnitude, including a planned $5 billion investment.^[1]

People and ownership

Meta tried to acquire SSI during the first half of 2025 and was rebuffed. In July 2025 Daniel Gross left the company for Meta Superintelligence Labs, and Sutskever became chief executive.^[1]

Where it sits in the field

SSI is an unusual shape of AI company. Its public record shows a very small team — roughly 20 people at the time of the March 2025 round — no revenue, and no released product, set against a valuation that rose from $5 billion to $30 billion in the space of about six months, and later a compute partnership with Nvidia that included a planned multi-billion-dollar investment.^[1] Most AI labs fund themselves by selling access to models; SSI's stated plan is to keep everything pointed at one target and to publish nothing in the meantime, which is what its "straight-shot" description is meant to capture.^[2]

That position also explains why the company is watched closely. The argument behind it is that safety cannot be bolted on to a superintelligent system after the fact, so the work on safety and the work on capability have to proceed as one problem rather than two.^[2] Whether that approach produces a usable result is an open question, and the company has not announced any findings, model or product so far.^[1] What exists publicly is the structure around the bet: a founding team assembled in June 2024, offices in California and Israel, large amounts of capital from venture investors, compute arrangements with Google Cloud and Nvidia, and a stated refusal to ship intermediate products.^[1][2]