Safe Superintelligence (SSI)

Safe Superintelligence Inc., known as SSI, is an artificial intelligence research lab with one of the most unusual business plans in the industry. It says it has one goal and one product: a safe superintelligence, meaning an AI system smarter than people that remains safe. It sells nothing. It has released no models to the public. Its website is a few short pages of plain text. That page says SSI is "an American company with offices in Palo Alto and Tel Aviv."

The lab was co-founded in June 2024 by Ilya Sutskever, one of the best known researchers in modern AI. This profile relies on SSI's own website and updates, its July 2026 announcement with NVIDIA and reporting summarized on Wikipedia and by Reuters. The company says very little in public, so this page says only what can be sourced.

What SSI says it is doing

SSI calls itself "the world's first straight-shot SSI lab." Its site lays out a few principles.

• One focus. The company says superintelligence is "within reach" and that building it safely is "the most important technical problem of our time."
• Safety and capability together. SSI says it will advance capabilities as fast as possible while keeping safety ahead, treating both as engineering problems.
• No product cycles. The company says its business model insulates safety, security and progress from short-term commercial pressure, with no management overhead or product launches to distract it.
• A small team. SSI says it is assembling "a lean, cracked team" of engineers and researchers. Wikipedia cites reports of about 20 employees in March 2025 and about 50 by July 2025.

These are the company's own claims. No outside party can check SSI's research progress, since it publishes little. We make no predictions about whether its goal is achievable or when.

The founders

Three people announced SSI on June 19, 2024.

• Ilya Sutskever co-founded OpenAI and served as its chief scientist. Earlier, as a graduate student at the University of Toronto, he worked with Geoffrey Hinton and Alex Krizhevsky on AlexNet, the 2012 image recognition system that set off the deep learning boom. He left OpenAI on May 15, 2024. He was also one of the OpenAI board members who voted in November 2023 to remove Sam Altman as chief executive, a decision reversed within days.
• Daniel Gross had led AI efforts at Apple and was a well known startup investor. With Nat Friedman he backed many AI companies through their fund, NFDG. One of them was Pika, the video startup on High Street; see our Pika profile.
• Daniel Levy is a former OpenAI researcher.

On July 3, 2025, SSI posted a message from Sutskever saying Gross was no longer part of the company as of June 29. Sutskever said he was now formally chief executive and Levy was president. Gross joined Meta's new superintelligence group, according to reporting at the time. Sutskever's note also addressed rumors that other companies wanted to buy SSI: "We are flattered by their attention but are focused on seeing our work through."

Funding and compute

SSI has announced little about its finances on its own site. What it has posted and what major outlets reported:

1. September 4, 2024. SSI said it had raised $1 billion from NFDG, Andreessen Horowitz, Sequoia, DST Global and SV Angel. Reuters reported a valuation of $5 billion.
2. Early 2025. Reuters reported in February 2025 that SSI was in talks to raise money at a valuation of at least $20 billion. Later 2025 reports described a round led by Greenoaks Capital at a valuation of about $30 billion or more. SSI did not publish those figures.
3. April 2025. Google Cloud agreed to provide its Tensor Processing Units for SSI's research, according to reporting. Terms were not disclosed.
4. July 26 and 27, 2026. SSI and NVIDIA announced a long-term strategic partnership that SSI says will let it scale its compute by ten times. The joint release names NVIDIA's Vera Rubin platform and says NVIDIA also invested in SSI. The release does not give the investment amount.

In the release Sutskever said SSI has research "worthy of scaling up" and that access to "a big NVIDIA computer" will let it do so. Our guide to AI investors on the Peninsula puts rounds like these in context.

SSI at a glance

The short list is itself telling. In more than two years SSI has posted only a handful of updates. That fits its stated plan to avoid product launches and publicity while it works.

Why Palo Alto

SSI's site says the company chose Palo Alto and Tel Aviv because the founders have "deep roots" in both places and can recruit top technical talent there. It does not list a street address. The NVIDIA announcement in July 2026 ran under a joint "Santa Clara, Calif. and Palo Alto, Calif." dateline, Santa Clara being NVIDIA's home.

The choice stands out because most of the largest AI labs settled in San Francisco. Palo Alto offers something different: a short trip to Stanford, to Sand Hill Road investors and to engineers who live on the Peninsula and would rather not commute to the city. Downtown Palo Alto has a long record of hosting small, intense research teams, from Google's first office at 165 University Avenue in 1999 to early Facebook. Our page on University Avenue's tech history covers those years. Our downtown guide describes the area today.

Stanford and the long view

Sutskever has a small Stanford connection. In 2012 he spent about two months as a postdoctoral researcher with Andrew Ng at Stanford before returning to Toronto to join Hinton's company, DNNResearch, which Google bought. The deeper link is intellectual. The term "artificial intelligence" was coined by John McCarthy, who later founded the Stanford Artificial Intelligence Laboratory in 1963. Our page on Stanford and the history of AI traces that history through SAIL, SRI's Shakey and the university's Institute for Human-Centered AI, which runs public talks on AI and its effects on society.

Stanford's HAI seminars are one of the few places a member of the public can hear researchers debate the questions SSI says it is working on. Our Stanford public lectures guide explains how to find them.

Visiting and practical notes

Frequently asked questions