Bezos's Prometheus raises $12bn to build an 'artificial general engineer'

Prometheus, the AI startup co-founded by Jeff Bezos, has raised 12 billion US dollars in a funding round that values the company at roughly 41 billion dollars. The deal comes only months after the firm emerged from stealth with a 6.2 billion dollar round late last year.

The company is pursuing what it calls an artificial general engineer, software intended to automate the design and manufacture of complex physical systems. According to the firm, that could span objects as varied as jet engines and pharmaceutical compounds, putting it squarely in the fast-growing physical-AI space rather than chatbots.

The ambition marks a deliberate departure from the crowded market for conversational AI. Where many of the best-known startups focus on generating text, images or code, Prometheus is positioning itself at the messier intersection of software and the physical world, where the work involves materials, tolerances, supply chains and the laws of physics rather than language alone.

Heavyweight backers and a small team

Bezos co-founded Prometheus with Vik Bajaj, a former co-founder of Verily, Alphabet's life-sciences unit, and the two serve as co-chief executives. Investors in the latest round include JPMorgan Chase, Goldman Sachs and BlackRock, alongside Bezos himself. Despite the eye-watering valuation, the company employs only around 150 people across San Francisco, London and Zurich.

The roster of backers is notable for its blue-chip character. The involvement of major Wall Street institutions, rather than purely venture-capital firms, signals that the appetite for large AI bets has spread well beyond Silicon Valley's traditional financiers. It also underscores how much capital investors believe is required to compete at the frontier of the field.

The contrast between the valuation and the headcount is striking. A company worth tens of billions of dollars with only around 150 employees reflects a pattern increasingly common in elite AI ventures, where small teams of specialists command enormous resources on the strength of their expertise and ambition rather than revenue or products already in the market.

Bezos plays down the job-loss fears

Speaking publicly about the venture for the first time, Bezos played down fears that such automation would destroy jobs, arguing the economy would face labour scarcity rather than mass unemployment. He suggested productivity gains would lift living standards and might let workers cut overtime rather than lose their roles entirely.

His comments tap into one of the central debates around advanced automation. Optimists argue that more productive economies create new kinds of work and raise prosperity overall, much as previous waves of technology did. Sceptics counter that the speed and breadth of AI-driven automation could be different in kind, displacing skilled engineering and design work that was once thought safe.

“Significant productivity in the economy is going to raise the standard of living.”

— Jeff Bezos, co-CEO of Prometheus

What the company says it is building

Details about Prometheus's technology remain scarce, but the broad pitch is clear. The firm wants to compress the painstaking cycle of designing, testing and manufacturing physical products, using AI to explore vast numbers of possible designs and predict how they will behave before anything is built. The potential targets it has named span several industries:

• Aerospace components such as jet engines
• Pharmaceutical compounds and drug design
• Complex industrial and manufactured systems
• Faster iteration between design and physical production
• Tools aimed at engineers rather than general consumers

If even part of that vision is realised, it could shorten development timelines that currently run for years in industries such as aerospace and pharmaceuticals. But applying AI to the physical world is far harder than to purely digital tasks, because mistakes can be costly, safety-critical and slow to verify.

Background: the rise of physical AI

Prometheus is part of a broader shift in the AI industry towards what is often called physical AI, the application of machine learning to robotics, manufacturing, materials science and engineering. Backers of the trend argue that the biggest economic gains from AI may ultimately come not from generating content but from transforming how things are made. The sheer scale of recent funding rounds in the area reflects that conviction.

For Bezos, the venture is also a notable personal turn. Having stepped back from day-to-day leadership at Amazon, he is now taking a hands-on co-CEO role at a company explicitly aimed at reshaping industry, a move that lends Prometheus both visibility and credibility regardless of how much of its technology has been proven.

“A multibillion-dollar valuation before a product ships is a bet on people and ambition as much as on technology.”

— Technology investor

What happens next

Much about the company's actual technology remains undisclosed, and a 41 billion dollar valuation for a barely public startup will fuel debate about froth in AI investing. The coming months should reveal whether Prometheus can move from ambition to demonstrable results, and whether its physical-AI thesis attracts imitators or scepticism. For now, the company has the capital and the names to make its ambitions hard to ignore.