Anthropic held acquisition talks this spring with Physical Intelligence, the robotics startup valued at $11 billion, according to reporting that followed a weekend rumor on X. Physical Intelligence CEO Karol Hausman denied a deal via a Slack message to employees containing a gif of a character from 'The Office' shaking her head no. That is not the world's most vigorous denial, and it leaves open the question of what actually transpired between the two companies earlier in the year.
Physical Intelligence has raised more than $1 billion since its founding roughly two years ago and was reportedly in talks this spring for a further $1 billion round at the $11 billion mark. Its π0.5 model is among the more widely used robot brains in academic robotics research. The company was co-founded by investor-operator Lachy Groom alongside former Google researchers and professors from Stanford and Berkeley.
Anthropic has made four known acquisitions this year. OpenAI has been more aggressive, buying at least 17 companies since 2023. Both filed confidentially for IPOs within a week of each other — Anthropic on June 1, 2026, OpenAI on June 8 — setting up what could be two of the largest US stock debuts in history and sharpening every strategic decision each company makes between now and pricing.
Key facts
- 01Anthropic and Physical Intelligence held acquisition talks this spring, per reporting, though the CEO denied a weekend rumor via a Slack gif.
- 02Physical Intelligence has raised more than $1B and was in talks for another $1B round at an $11B valuation.
- 03Anthropic has made 4 known acquisitions in 2026; OpenAI has bought at least 17 companies since 2023.
- 04Anthropic filed confidentially for IPO on June 1, 2026; OpenAI filed a week later on June 8.
- 05OpenAI is already an investor in Physical Intelligence, sharing backers Khosla Ventures, Thrive Capital, and Founders Fund.
The robotics question is the interesting one. Physical-world understanding may be a prerequisite for the systems both labs say they want to build, and no volume of internet text substitutes for it. Buying an existing robotics team lets a lab skip years of hardware iteration and dataset collection.
OpenAI's own history illustrates the difficulty. The company built an early robotic hand that solved a Rubik's Cube, then shut its entire robotics group down in 2021. Co-founder Wojciech Zaremba later explained the decision.
“the approach was missing pieces needed for real superintelligence”— Wojciech Zaremba, OpenAI co-founder
The team quietly returned in 2024, building a humanoid robotics lab in San Francisco. Sam Altman made 'OpenAI Robotics' official in late May, describing a near-term focus on robots for infrastructure work and a long-term goal of a personal robot for everyone. Hiring is underway.
Anthropic has taken a different path. Rather than building hardware, it has run robotics through its safety-research group. Project Fetch, announced last November, tested how much Claude could help non-experts program a robot dog. A second phase in June found that a newer Claude model completed the same tasks roughly 20 times faster than the best human-plus-Claude team from the year before. The results are impressive on paper but do not amount to a robotics business, which is why an acquisition of a company like Physical Intelligence would make sense.
There is a complication, and it is a significant one. Physical Intelligence's early investor base looks a lot like OpenAI's own. Khosla Ventures and Thrive Capital are on both cap tables. Founders Fund, another major OpenAI backer, was reportedly involved in Physical Intelligence's newest round earlier this year. And OpenAI itself is an investor in Physical Intelligence.
That last fact matters. OpenAI is not a peripheral observer here — it is a shareholder in a company its chief rival was reportedly negotiating to buy. Large strategic investors sometimes negotiate information rights or rights of first refusal over sales to competitors, precisely for this scenario. Whether OpenAI holds such provisions in Physical Intelligence is unclear, and the company did not respond to questions about it.
Physical Intelligence's denial is worth taking at face value in the narrow sense — no deal was signed — but the underlying dynamic is what matters. A company with a leading robotics model, deep ties to Silicon Valley's most powerful investors, and an $11 billion valuation is being circled by the two labs racing to define what a general-purpose AI system looks like. That the CEO's denial came in gif form rather than a categorical statement is telling.
If Physical Intelligence is actually in play, OpenAI has the more obvious claim on paper — already a shareholder, already close to Groom, already building a robotics organization that would absorb the team more cleanly. For Anthropic, coming from behind on hardware, the strategic logic of an acquisition is stronger but the execution is harder, especially with a competitor holding equity. Either way, the weekend rumor exposed how much of the next phase of the AI race will be settled not by benchmark scores but by which companies can staple physical-world capability onto their models fastest — and who they have to outbid, or out-negotiate, to do it.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
