Safe Superintelligence company profile: valuation, funding & revenue

This Safe Superintelligence company profile covers the company's valuation, revenue, funding rounds, investors and history. Safe Superintelligence works in artificial intelligence, with a focus on superintelligence research. It was founded in 2024 and is headquartered in Palo Alto, California, United States (with an office in Tel Aviv). Its latest valuation on record is US$32 billion (April 2025). Every figure links to a public source, and anything the company hasn't disclosed is marked as such.

Revenue

Funding history

Safe Superintelligence timeline

Key dates from the sources on this page, oldest first. Metric snapshots are listed after the funding events.

• 2024 — Safe Superintelligence is founded in Palo Alto by Ilya Sutskever (CEO), Daniel Gross and Daniel Levy.
• Sep 2024 — Equity round: US$1B, valuing the company at US$5B.
• Apr 2025 — Equity round: US$2B, valuing the company at US$32B. Investors: Greenoaks (lead); Andreessen Horowitz, Lightspeed, DST Global; strategic: Alphabet, Nvidia.
• April 2025 — Compute: Trains on Google's TPUs and was reported to be Google Cloud's largest external TPU customer.

Investors

Overview

Safe Superintelligence Inc., usually called SSI, is an AI research company with a single stated goal: to build "safe superintelligence", AI that surpasses human intelligence while remaining safe. Unlike most AI companies, it has said it will not release products along the way. It doesn't sell a chatbot, an API or enterprise software. Its entire focus is research towards that one objective.

SSI was founded in June 2024 by Ilya Sutskever, Daniel Gross and Daniel Levy. Sutskever is one of the most influential researchers in modern AI: he co-authored the 2012 AlexNet paper that launched the deep learning era, co-founded OpenAI and served as its chief scientist, and was among the board members who briefly removed Sam Altman as CEO in November 2023 before reversing course. Gross had led AI efforts at Apple and was an investor; Levy was a researcher at OpenAI. The company has offices in Palo Alto and Tel Aviv. Gross left in mid-2025, and Sutskever took over as CEO.

A deliberately different strategy

Most AI labs pay for research by selling products, which also gives them real-world feedback. SSI has argued that commercial pressure and product deadlines distract from the hard problem of building safe superintelligence, so it has insulated its team from that pressure. Its small team, reported at around 20 people in 2025 and a few dozen later, works on research it has largely kept private. Sutskever has suggested in interviews that the industry's focus on simply scaling up existing methods is reaching its limits and that new research ideas are needed.

Funding

• September 2024: about US$1 billion at a US$5 billion valuation, three months after founding.
• April 2025: US$2 billion at a US$32 billion valuation, led by Greenoaks with a reported US$500 million, with Andreessen Horowitz, Lightspeed Venture Partners and DST Global. Alphabet and Nvidia joined as strategic backers.

Later reports in 2026 described a large Nvidia investment linked to a computing partnership. Those terms were not disclosed by the companies, so we don't include them in our data. SSI's valuation rose more than six times in seven months without a product, which shows how much investors are willing to pay for a small number of exceptional researchers.

Revenue

SSI has no products and no revenue. That is a deliberate choice, not an oversight. It also means its value depends entirely on its research succeeding, and on investors continuing to fund it until it does.

Compute

Training advanced AI requires enormous computing power. SSI trains on Google's Tensor Processing Units (TPUs) rather than relying only on Nvidia GPUs and was reported in 2025 to be Google Cloud's largest external TPU customer. Its later relationship with Nvidia suggests it is diversifying its access to chips, a common strategy as labs compete for scarce capacity.

Competitors

SSI competes for talent and attention with OpenAI, Anthropic, Google DeepMind, Meta, xAI and Thinking Machines Lab. It doesn't compete with them for customers, since it has none, but it competes intensely for researchers and computing power, both scarce.

SSI in brief

Safe Superintelligence is an AI research company founded in June 2024 by Ilya Sutskever, Daniel Gross and Daniel Levy, based in Palo Alto and Tel Aviv. It raised about US$1 billion at US$5 billion in 2024 and US$2 billion at US$32 billion in April 2025, led by Greenoaks, with Alphabet and Nvidia as strategic investors. It has no products or revenue by design.

How investors value a lab with no product

Traditional valuation methods rely on revenue, profit or at least user growth. SSI has none of these. Investors are instead making an option-style bet: a small chance that SSI produces breakthroughs leading to enormously valuable technology, weighed against a high chance that it doesn't. The reputation of the founding researchers carries most of the weight, along with the scarcity of people capable of frontier AI research. This kind of valuation can rise or fall dramatically on technical progress, departures or a change in market sentiment towards AI. Our guide to pre-revenue valuation methods covers the approaches investors use in similar cases.

Safety as a business principle

SSI's name puts safety at the centre of its identity. In practice, AI safety covers several things: making sure systems do what their designers intend, preventing misuse, and avoiding catastrophic failures as systems become more capable. SSI has said it intends to advance safety and capabilities together, keeping safety ahead. Because it publishes little, outsiders can't easily judge its approach, which is one of the main criticisms of its secrecy.

Israel's role

SSI's Tel Aviv office reflects Sutskever's ties to Israel, where he grew up, and the country's deep pool of technical talent. Israel has become an important location for AI research, with offices of major labs and many startups. For SSI, it offers access to researchers outside the intensely competitive Silicon Valley market.

For now, most of what the outside world knows about SSI's work comes from occasional interviews with its founders rather than from any published research papers.

What could change the story

Any of the following would be significant: a research publication or demonstration, a confirmed new funding round, a change in its no-product policy, senior departures or hires, or a large computing deal.

For different readers

AI researchers see SSI as one of the few labs focused purely on long-term research. Founders can note how far reputation alone can carry a fundraise in frontier AI. Investors should treat any exposure as a high-risk, long-term bet with no near-term revenue.

Latest developments (October 2026)

Safe Superintelligence remains secretive about its research. In July 2026 AI Magazine reported a partnership with Nvidia that would give the lab access to Nvidia's newest Vera Rubin systems, with a reported investment; we have not seen the size confirmed by the companies. Our profile keeps the US$32 billion valuation from its 2025 round. The company has not released a product. See our new AI labs guide.

Sources

Every figure on this page comes from the public sources below. Where sources disagreed, we used the company's own announcement or a primary news report. Private-company data changes; check the date beside each figure.

1. Calcalist
2. Wikipedia