This year has been a big year for AI company acquisitions. It’s so big that most companies are barely registered anymore. Anthropic and OpenAI have each acquired developer tools, AI service shops, and product testing startups, and each continues to do so in order to convert model capabilities into enterprise revenue and expand their reach faster than the other. That’s what made rumors over the weekend that Anthropic would acquire robotics startup Physical Intelligence stand out. Even after Physical Intelligence’s CEO denied it, the virus spread very quickly.
Part of it has to do with who is involved. Physical Intelligence is not some obscure robotics store. The company was co-founded by Lachy Groom, an investor and executive who has gained attention in Silicon Valley in recent years. The company has raised more than $1 billion (and was reportedly in talks to raise another $1 billion round this spring at an $11 billion valuation). And its π0.5 model seems to be one of the more widely used robot brains in robotics research.
As it turns out, the rumors weren’t completely false. According to The Information, Anthropic and Physical Intelligence actually held acquisition talks this spring, so it’s possible that technology blogger Robert Scoble (whose weekend post about X sparked the frenzy) misunderstood the details without giving the impression that something had happened.
It’s worth noting that Physical Intelligence’s reaction to the rumors wasn’t the world’s most aggressive denial. According to The Information, Karol Hausmann, CEO of Physical Intelligence, told employees the report was untrue via a Slack message that included a gif of a character from “The Office” shaking his head.
Groom did not respond to TechCrunch’s request for comment sent Monday night.
Anthropic has made four known acquisitions this year. OpenAI has become more aggressive, acquiring at least 17 companies since 2023. Of course, both companies are preparing to go public. Anthropic secretly filed for an IPO on June 1, followed by OpenAI a week later, setting up two stocks for what could be the biggest U.S. stock debut in history.
So why embrace robotics now? The most likely answer is that understanding the physical world may be a prerequisite for superintelligent systems, and no amount of internet text can replace it.
OpenAI’s own history here is instructive. The company developed an early robotic hand that could solve a Rubik’s Cube, but then shut down its entire robotics group in 2021, with co-founder Wojciech Zaremba later saying the approach lacked the elements needed for true superintelligence. The team returned in 2024 and was quietly building a humanoid robot lab in San Francisco, but in late May CEO Sam Altman officially announced the company’s founding, with OpenAI Robotics announcing the hiring and explaining a short-term focus on robots for infrastructure work and a long-term goal of providing personal robots for everyone.
Anthropic isn’t building anything like OpenAI’s hardware lab. What the company did was publish a series of research results through an internal group that stress-tested Frontier’s capabilities for safety purposes. That included last November’s Project Fetch, where Anthropic staff tested how well Claude could help non-experts program a robot dog. According to Anthropic, the second phase in June found that the new model completed the same tasks approximately 20 times faster than the best human-Claude team from the previous year.
Acquiring an existing team with robotics expertise could save Anthropic years of work. However, complications can arise. Physical Intelligence was founded in San Francisco about two years ago by Groom, former Google researchers, and professors from Stanford and Berkeley. Its early investor base is very similar to OpenAI’s, including Khosla Ventures and Thrive Capital. Founders Fund, which is also a major investor in OpenAI, reportedly participated in Physical Intelligence’s latest funding round earlier this year.
In fact, OpenAI itself is investing in physical intelligence, so it’s more than just a peripheral. That’s a shareholder in a company whose biggest rival has reportedly been in takeover talks very recently.
This raises the question of whether OpenAI’s initial investment came with any information rights or rights with a right of first refusal against sale to a competitor, the kind of protection clauses that strategic investors sometimes negotiate for precisely this scenario.
So if physical intelligence is indeed at work, it remains possible that OpenAI (which is already a shareholder, already close to Groom, and already trying to ensure it surpasses Anthropic in robotics) has a clearer claim than Anthropic. We asked OpenAI these questions earlier today, but the company did not respond.
If you buy through links in our articles, we may earn a small commission. This does not affect editorial independence.
