
- Prometheus, the AI startup co-led by Jeff Bezos and Vik Bajaj, raised $12 billion in Series B funding at a roughly $41 billion valuation — just seven months after launching with a $6.2 billion Series A.
- The investor list reads like a Wall Street roll call: JPMorgan, BlackRock, Goldman Sachs, DST Global and Arch Venture Partners, with Bezos himself participating again.
- The company is building an “artificial general engineer” — AI tools meant to make the design-to-manufacturing cycle for physical products 10x faster or more, from jet engines to chips.
- With about 150 employees and no product timeline, Prometheus is already worth more than most public industrial companies — and a reported $100 billion affiliated fund may be next.
Here is a thought experiment for anyone who believes AI valuations have peaked. A startup with roughly 150 employees, no shipped product, no announced timeline, and a training-data problem its own founders admit is unsolved just convinced JPMorgan, BlackRock and Goldman Sachs to help write a $12 billion check. The company is Prometheus, the price tag is a $41 billion valuation, and the co-CEO is Jeff Bezos — in his first chief executive role since leaving Amazon in 2021. On June 11, Bezos and co-CEO Vik Bajaj gave their first joint interview and explained what the money is actually for. The answer says a lot about where the smartest capital thinks AI goes next: not chatbots, but factories.
A $41 Billion Bet on the Physical World
From $6.2B Series A to $12B Series B in seven months
Prometheus launched in November 2025 with a $6.2 billion Series A — at the time one of the largest first rounds ever raised — with Bezos as the largest backer. The new $12 billion Series B, announced June 11, values the company at roughly $41 billion and brings in JPMorgan, BlackRock, Goldman Sachs, DST Global and Arch Venture Partners. Bezos participated in this round as well. Unlike Blue Origin, which Bezos has famously bankrolled alone by selling Amazon stock, Prometheus is structured to absorb outside institutional capital at scale. “This is a capital-intensive startup, there’s no question about that,” Bezos told CNBC, citing the cost of compute and of building the specialized training data the company needs. Asked about an IPO, he said it is “too early to think about that.”
The company has also quietly rebranded — dropping the “Project” from “Project Prometheus” — and now operates with around 150 employees across San Francisco, London and Zurich. That works out to roughly $273 million of valuation per employee, a ratio that exceeds even the frothiest AI lab benchmarks.
Business Insight — The investor mix is the tell. JPMorgan, BlackRock and Goldman are not typical Series B leads — they are balance-sheet institutions that show up when they expect an asset to become a permanent fixture of the capital markets. Their presence signals that Wall Street is now underwriting industrial AI as an asset class, not a venture experiment.
What an “Artificial General Engineer” Actually Does
The jet engine problem
Bezos gave Axios the clearest articulation yet of the product thesis. “If you go to a current jet engine manufacturer and say you want the exact same engine but with 10% more thrust, it could be a 10-year program,” he said. “Not because they’re lazy or bad at their jobs, but because it’s so complex. So what we’re doing is building a set of tools that will empower engineers to compress that cycle time and make that dream-build loop be 10 times faster or even more.” Bajaj — a Stanford-affiliated scientist who co-founded Alphabet’s Verily — added that what changed in the last few years is the ability to formulate something as complicated as a jet engine, from design to manufacturing, “as an end-to-end AI problem.”
Crucially, Prometheus is not a robotics company and is not trying to automate factories, a point Bezos has stressed repeatedly. The focus is on pre-production: design, prototyping, simulation and process optimization — the slow, expensive iteration loop that precedes manufacturing at scale. The company has no corporate ties to Amazon or Blue Origin, though Bezos called Blue Origin “a case study for a customer of Prometheus.” Both founders insist the technology will create more engineering jobs, not fewer. “The pace of our physical creation right now is nowhere near the pace of human imagination,” Bajaj argued.
Business Insight — Software AI markets are crowded and margin-compressed; the physical economy is not. Aerospace, medical devices, chips and heavy machinery represent trillions in annual output where a 10x faster design loop converts directly into market share. If Prometheus works, its customers’ R&D budgets — not consumer subscriptions — become the revenue pool, which is exactly the kind of moat-rich, enterprise-locked market Bezos built AWS on.
The $100 Billion Question Nobody Will Answer
No “Internet of manufacturing data” exists — so Bezos may buy one
The most revealing part of the announcement is what the co-CEOs declined to discuss. There is a reported effort to raise as much as $100 billion for an affiliated holding company that would acquire legacy industrial businesses — companies whose proprietary manufacturing data would feed Prometheus’ models, and which would in turn receive its AI tooling. Bezos would say only that Prometheus may buy parts of companies that could benefit from its technology. The founders also acknowledged a structural problem: unlike language models trained on the open web, there is no “Internet of manufacturing data” to ingest. Decades of engineering know-how sits locked inside corporate archives, CAD systems and the heads of retiring specialists. How Prometheus is being trained, and when its first product ships, remain undisclosed — the co-CEOs said only that early rollouts are coming.
Why speak publicly now? Bezos was blunt: the story was “kind of leaking out,” and “if you let that just be a complete void, they’ll fill it with nonsense.” The interview also let him address Blue Origin’s New Glenn launchpad explosion last month — “a very bad day for Blue” — while confirming the rocket will fly again before year-end.
Business Insight — The data scarcity problem is the real story. If the only way to train an industrial AI is to own the factories that generate the data, then the $100 billion acquisition fund is not a side project — it is the strategy. That would make Prometheus less like an AI lab and more like a private-equity roll-up with a model attached, a playbook incumbents like Siemens and GE will struggle to copy at this capital scale.
Related
- This AI Quietly Captured 40% of the Fortune 50
- Accenture Just Bet $110M to Replace the Call Center
- Amazon Just Quietly Came for Redbubble’s Whole Business
- Perplexity Let AI Write Its Own Search — And It Won
Sources
- GeekWire — Bezos’ AI startup Prometheus raises $12B at $41B valuation, and the CEOs explain what they’re doing (2026-06-11)
- Axios — Prometheus, Jeff Bezos’ AI startup, is now worth $41 billion (2026-06-11)
- CNBC — Bezos opens up about AI startup Prometheus after $12 billion raise (2026-06-11)
AI Biz Insider · AI Business EN · aibizinsider.com
댓글 남기기