It’s been a big year for AI adoption – so big that most of them don’t even register anymore. Anthropic and OpenAI have each gone on buying sprees, using developer tools, AI resource stores, and product testing startups to turn the power of the model into business revenue and expand their reach faster than the other. What made the weekend rumors about Anthropic’s start-up finding robots Physical Intelligence stand out. It spread very quickly, even after being denied by the CEO of Physical Intelligence.
Part of that has to do with who is involved. Physical Intelligence is not some vague robot shop. It was co-founded by Lachy Groom, an investor-operator whose star has been rising in Silicon Valley in recent years; has raised more than $1 billion in funding (and was reportedly in talks this spring for another $1 billion round at an $11 billion valuation); and its π0.5 model is apparently among the most widely used robot brains in robotics research.
As it turned out, the rumor was not entirely true. Anthropic and Physical Intelligence actually held fundraising talks this spring, according to The Information, so tech writer Robert Scoble — whose weekend X post caused quite a stir — may have gotten the wrong information without being mistaken for something.
Physical Intelligence’s response to the rumor was not the strongest denial in the world, it should be noted. According to a report by The Information, Physical Intelligence chief Karol Hausman told employees that reports are false about a Slack message containing a gif of a character from “The Office” shaking a number.
The groom, meanwhile, did not respond to TechCrunch’s request for comment, which it sent Monday night.
Anthropic has done four notable things this year; OpenAI has been very aggressive, acquiring at least 17 companies since 2023. Both of them, of course, are now preparing to go public. Anthropic filed privately for an IPO on June 1, followed by OpenAI a week later, setting off what could be the two largest US stock IPOs in history.
So why robots, why now? The most likely answer is that global understanding may be a requirement for the most intelligent systems, and no amount of internet text can replace it.
The history of OpenAI here is instructive. It built the first robotic hand that could solve Rubik’s Cube, then shut down the entire robotics team in 2021, when founder Wojciech Zaremba later claimed the method lacked the necessary pieces of real intelligence. The team returned in 2024, quietly building a lab of humanoid robots in San Francisco, before CEO Sam Altman made it official in late May, announcing that “OpenAI Robotics” was hiring and explaining the close focus of robots on infrastructure work, with a personal robot for everyone as a long-term goal.
Anthropic has yet to build anything like an OpenAI hardware lab. What it has done is publish a series of research pieces on its internal group that examines border forces for security purposes. That included Project Fetch last November, where Anthropic staff tested how much Claude could help non-professionals program a robot dog, and the second phase in June that, according to Anthropic, we got a new model that completed the same tasks about 20 times faster than the best team of people-plus-Claude from last year.
Buying an existing team with robotics technology will allow Anthropic to skip years of operations. There is a potential problem, however. Physical Intelligence was founded in San Francisco about two years ago by Groom, former Google researchers, and Stanford and Berkeley professors, and its early investor base appears to be similar to OpenAI’s, including Khosla Ventures and Thrive Capital. Founders Fund – which is also a major investor in OpenAI – reportedly participated in a new funding round for Physical Intelligence earlier this year.
In fact, OpenAI itself is an investor in Physical Intelligence, so it’s not just a side player; is another shareholder in the company whose biggest rival is reported to be buying it recently.
That raises questions about whether OpenAI’s early investment came with any intellectual property rights, or the right of first refusal to sell to a competitor — the kind of protections strategic investors sometimes negotiate to get this position.
That leaves open the possibility that if Physical Intelligence really plays out, OpenAI — already a shareholder, already close to Husband, already trying to ensure Anthropic’s lead in robotics — could have a more obvious claim than Anthropic. We asked OpenAI these questions earlier today and the company did not respond.
If you shop through links in our articles, we may earn a small commission. This does not affect our editorial independence.
