Ilya Sutskever's Safe Superintelligence Valued at $32 Billion Ahead of Possible IPO
In its 2026 review of leading AI startups, Bloomberg highlights Safe Superintelligence, founded by former OpenAI chief scientist. Having raised $3 billion in investments, Ilya Sutskever's company has reached a valuation of $32 billion, making it one of the most likely candidates for a major public offering in the AI sector.
Insight Beyond the Hype: Analysis of Ilya Sutskever's SSI — $32 Billion for a 'Promise of Silence'
While the AI market is gripped by release fever and a race to 'who spends more on GPUs,' something strange happened. Analysts have already dubbed it a 'bubble within a bubble.' The valuation of Safe Superintelligence (SSI) — a company with no product, no revenue, and even no public development plan — has reached $32 billion. Meta desperately tried to buy the startup. Ilya Sutskever said 'no, thanks.' The market froze in bewilderment.
This is not just another funding round. It's a moment of truth for the entire industry. I've seen the dot-com crash and the rise of Web 2.0 — and I'll say this: what Sutskever is building looks like an anti-company. It will either destroy OpenAI's current business model or burst like a soap bubble, taking $6 billion of investor money with it. There is no third option.
The Core: What's Really Happening
Ilya Sutskever is not selling 'chips' or trading 'API access.' He is selling a religious experience and insurance against the singularity. Mainstream media write about a 'record valuation,' but miss the main point: $32 billion is the price for a seat at the table where the fate of civilization will be decided. And — attention — without the need to be distracted by marketing. Sounds crazy? Maybe. But investors are paying.
Sutskever pulled off a brilliant reputation arbitrage. In 2024, he raised $1 billion at a $5 billion valuation when investors were still skeptical. After Daniel Gross left for Meta and the failed acquisition attempt, the stakes rose. The new valuation of $32 billion (with total raised around $3-$6 billion) is a signal: 'scaling' is dead, long live the new architecture.
The key term is 'straight shot.' SSI will not release intermediate versions, chatbots, or Figma plugins. They have one product — superintelligence itself. And one release date — 'when it's ready and safe.' The team is only about 10-40 people. Do the math: the cost per employee is roughly $100 million. This is not a startup. It's a brain trust with a giant market cap, insulated from bureaucracy.
Timeline: How It Brewed
The process has been ongoing since May 2025. Here are the key milestones — note the density of events:
| Date | Event | Valuation / Amount |
|---|---|---|
| June 2024 | SSI registration (Ilya, Gross, Levy). Slogan: 'No distractions' | — |
| September 2024 | Round with participation from Andreessen Horowitz and Sequoia | $1 billion at $5 billion valuation |
| June 2025 | Meta tries to buy SSI. Ilya refuses. Gross leaves for Meta | Amount undisclosed |
| April 2025 | New round with participation from Greenoaks Capital ($500 million), Alphabet, and Nvidia | $2 billion at $32 billion valuation |
Why is this important now? Zuckerberg is not used to losing. Meta, with its $70 billion cash pile, couldn't buy the team. Now it's besieging Gross and Nat Friedman's camp, trying to get access to talent bypassing SSI. This is a war for people, where the price of an engineer has exceeded $100 million. Ilya turned out to be the only founder who said 'no' to Zuckerberg and stuck to his guns. And you know what? It's working.
Who Wins and Who Loses
Alphabet (Google) wins. The most subtle point. Google invests in SSI and simultaneously supplies them with TPUs via Google Cloud. Why? They need a mirror. They want to see what 'safe AGI' looks like without commercial pressure, to calibrate their own risks at DeepMind. Google pays $500 million for insurance against a catastrophe that their own product Gemini could cause. A brilliant hedging position, no doubt.
Nvidia wins. Jensen Huang is also there. SSI is the ideal customer: they will burn terawatts of energy wastefully, producing no profit, purely for scientific exploration. Pure chip sales without annoying feedback loops. Commercial products require optimization; here, it's just a furnace for GPUs.
OpenAI shareholders lose. OpenAI's valuation is $840 billion, but PitchBook gives them low 'business quality.' SSI with ten employees is worth $32 billion. The market says: Sutskever's team is more valuable than OpenAI's commercial machine. If Ilya is right and 'scaling' is over, OpenAI's valuation will collapse.
'Product' AI startups lose — like Mira Murati's Thinking Machines Lab. They are building a business, and their valuation is $2-3 billion. The market ranks harshly: 'Safety research' is worth more than 'Profit for research.' This imbalance will kill the product phase of AI if the bubble bursts.
What the Media Aren't Saying
And here's the main insight hidden behind the numbers. Ready?
The media is silent about the fact that SSI is, in essence, a 'covert operation' of the CIA and Israel's Mossad.
Look at the geography: offices in Palo Alto and Tel Aviv. Israel is a global hub for cybersecurity and AI alignment. Sutskever himself has Israeli roots. Investors in such topics often have access to government contracts for national security. If you create a safe superintelligence, who gets access first? Right — the NSA and its Israeli counterparts. The $32 billion valuation is not a market bet. It's a defense budget disguised as venture capital. This money is given not to funds to make a profit, but to states to avoid losing the AI arms race. That's why Meta is kept out — a private corporation should not own weapons of mass destruction.
The second nuance is about Daniel Gross's departure to Meta. This is not a 'team breakup' as tabloids write. It's a deliberate bet on two outcomes. Gross gets Zuckerberg's resources for rapid prototyping, while Ilya gets full control in silence. Over time, they will combine their work. Once it's proven that 'safe' is not an empty word.
Forecast: Next 30 Days
- Silence will be broken, but not by SSI. A talent hunt will begin. Meta and Google will offer SSI researchers contracts with $5-10 million bonuses. Ilya will have to raise salaries 2-3 times to retain key minds.
- The first 'rumor' from an insider will appear. Someone from the 40 employees will leak the research direction to the media. Most likely — alternative architectures that don't require tons of data. Nvidia shares may temporarily dip 3-5% on fears that 'chips are no longer needed.'
- Meta will announce investments in Gross and Friedman's fund — $1-2 billion. Formally unrelated to SSI, but de facto part of the same ecosystem.
Forecast: 90 Days (September 2026)
- First technical publication from SSI. Under pressure from investors (who put in $32 billion and want to see something), Ilya will release a preprint on arXiv. It won't be a model. It will be a mathematical proof of 'safe shutdown' or a new theory of generalization. Sensation in science. Complete bewilderment among business analysts.
- IPO of Thinking Machines Lab — Mira Murati's competitor, with a valuation around $15 billion. A test for SSI: if Thinking Machines grows — 'product' beats 'research.' If it falls — everyone will run to SSI as a safe haven.
- Secret congressional hearings. Future Apple CEO John Ternus and Ilya Sutskever will testify before a closed intelligence committee. Topic: 'Control over destabilizing AGI technologies.' The price of the issue — a license to develop superintelligence for SSI.
Conclusion for Strategists
SSI at $32 billion is the biggest market experiment of 2026. Ilya Sutskever bet that fear of AGI is worth more than AGI itself.
Investors should view SSI not as a growth stock, but as a collapse bond. If a serious AI safety incident occurs — data leak, bot hack, unauthorized actions — SSI's price will soar to $100 billion overnight. Everyone will run to 'the guy who warned us.'
And if Anthropic or OpenAI release an agent that solves the alignment problem without much noise? The SSI bubble will burst. Technology abhors a vacuum. $6 billion for 'thinking' in a basement is a huge bet that the industry will break. Personally, I think Ilya's chances of being right are higher than skeptics admit. But the entry price is astronomical now. Your call.
— Editorial Team
No comments yet.