FieldAI Secures $700M at $10B Valuation to Build a Universal Robot Brain

The race to build a general-purpose brain for robots has a new giant. Robotics startup FieldAI has reportedly secured $700 million in new funding at a $10 billion valuation, according to a report from Business Insider, in one of the largest physical-AI financings of the year.

The company is said to have signed a term sheet for the deal, though the lead investor was not disclosed.

Founded in 2023 in Irvine, California, FieldAI develops Field Foundation Models — AI software that lets robots navigate and work on their own without maps, GPS, an internet connection, or pre-defined travel paths.

The $10 billion valuation is roughly five times its August 2025 valuation of around $2 billion. The round puts FieldAI in the top tier of the physical-AI race, alongside heavily funded rivals such as Physical Intelligence, valued near $11 billion, and Skild AI, valued above $14 billion.

All three aim to become the software layer powering humanoid robots, robot dogs, drones, and industrial machines.

Most industrial robots today rely on painstaking environmental mapping, GPS, or constant connectivity. FieldAI's models are designed to work where none of those exist.

The software fuses data from cameras, lidar, and radar into a continuously updated digital twin — a live simulation of the robot's surroundings — and automatically adjusts behavior in response to risk, such as slowing down in low-visibility areas.

Customers can also use the digital twins to test site changes virtually before altering the real workspace.

The approach has gained traction in industries where environments change daily. In March 2026, FieldAI partnered with Boston Dynamics to deploy its technology on the company's four-legged Spot robot for industrial equipment inspection at construction sites, where shifting layouts make static maps nearly useless.

The company says its software runs across everything from quadrupeds and humanoids to wheeled robots and passenger-scale vehicles.

What sets FieldAI apart from many foundation-model startups is a real revenue number. CEO Ali Agha, a former NASA Jet Propulsion Laboratory engineer who spent seven years leading autonomy work on the DARPA Subterranean Challenge, told Business Insider the company has seen rapid growth in recent months.

Since June, FieldAI has added at least $35 million in new revenue and customer contracts, bringing its total past $135 million across more than 30 customers in construction, energy, the public sector, defense, and data centers.

The company's backers read like a who's who of tech investing: Bezos Expeditions, Nvidia's NVentures fund, Intel Capital, Khosla Ventures, and Laurene Powell Jobs' Emerson Collective. It has also been hiring engineers from Google DeepMind, Tesla, Nvidia, and Boston Dynamics as the competition for robotics talent intensifies.

The valuation reflects a bet that the most valuable layer in robotics is software, not hardware.

Rather than building a single robot, FieldAI aims to sell one universal general-purpose brain that runs across many different machines and industries — a pitch that becomes more attractive as robots move from controlled demonstrations into messy real-world settings like construction sites, mines, and factory floors.

The $135 million figure combines revenue with signed contracts, but the growth from over $100 million in June has been enough to justify a 5x valuation jump in just over a year.

If the round closes as reported, FieldAI will have deep pockets to fund its expansion into manufacturing and defense deployments while the rest of the industry races to keep up.