• 3 min read
Anthropic’s robotics talks with Physical Intelligence surfaced
A viral rumor that Anthropic was buying Physical Intelligence was overstated, but the companies did discuss an acquisition this spring.
Image: TechCrunch
A weekend rumor that Anthropic was acquiring robotics startup Physical Intelligence spread quickly across X, even after a denial from the startup’s CEO. According to The Information, the story had a real basis: Anthropic and Physical Intelligence did hold acquisition talks this spring.
That helps explain why the claim took off. Physical Intelligence is a high-profile robotics company, co-founded by Lachy Groom, that has raised more than $1 billion. It was also reportedly in talks this spring for another $1 billion round at an $11 billion valuation. Its π0.5 model is said to be one of the more widely used robot foundation models in robotics research.
The denial itself appears to have left room for doubt. Per The Information, CEO Karol Hausman told employees the reports were false in a Slack message that included a gif from “The Office” of a character shaking her head no. Groom did not respond to TechCrunch’s request for comment Monday night.
Recommended reading
Kimi K3 Ranks No. 2 on AA-Briefcase
The timing matters. Anthropic has made four known acquisitions this year, while OpenAI has bought at least 17 companies since 2023. Both companies have also filed confidentially to go public: Anthropic on June 1, and OpenAI a week later.
Why robotics is suddenly strategic
One reason for the interest is simple: training systems to understand the physical world may be critical to building more capable AI. OpenAI explored that earlier with a robotic hand that solved a Rubik’s Cube, then shut down its robotics group in 2021. It restarted those efforts in 2024 with a humanoid robotics lab in San Francisco, and in late May, CEO Sam Altman publicly said OpenAI Robotics was hiring, with an initial focus on infrastructure work and a long-term goal of personal robots.
Anthropic has taken a different route. Rather than building a hardware lab, it has published research through its internal safety group, including Project Fetch last November. In that work, Anthropic tested how well Claude could help non-experts program a robot dog. In a second phase published in June, the company said a newer model completed the same tasks about 20 times faster than the best human-plus-Claude team from the prior year.
Acquiring a robotics team would let Anthropic move faster. But OpenAI may have an inside track.
OpenAI’s existing stake could matter
Physical Intelligence was founded in San Francisco about two years ago by Groom, former Google researchers, and professors from Stanford and Berkeley. Its backers include Khosla Ventures and Thrive Capital, both also tied to OpenAI. Founders Fund, another major OpenAI investor, was reportedly involved in Physical Intelligence’s latest round earlier this year.
Most notably, OpenAI is itself an investor in Physical Intelligence. That raises unanswered questions about whether OpenAI’s investment includes information rights or a right of first refusal in the event of a sale to a competitor. TechCrunch asked OpenAI about that possibility and did not receive a response.
If Physical Intelligence is actually in play, that existing relationship could make OpenAI — not Anthropic — the more natural buyer.
AI Editor
Ava covers the rapidly evolving world of artificial intelligence, from foundational models and research labs to the real-world economics of intelligence. With a background in computational linguistics, she cuts through the hype to find out what actually works. She firmly believes that benchmarks are just marketing until reproduced in the wild.
via TechCrunch
