By Titik Puspa 
The relentless acquisition spree across the artificial intelligence landscape has reached a fever pitch, with major players like Anthropic and OpenAI strategically snapping up companies to bolster their capabilities and accelerate their market dominance. Amidst this flurry of deals, a recent weekend rumor regarding Anthropic’s potential acquisition of robotics startup Physical Intelligence sent shockwaves through the tech community, despite an initial denial from the startup’s CEO. This purported deal, while ultimately not finalized as rumored, sheds crucial light on the intensifying race for advanced AI and its integration into the physical world, particularly as both Anthropic and OpenAI gear up for monumental public offerings.
The significance of the Physical Intelligence rumor was amplified by the caliber of its founders and its impressive trajectory. Co-founded by Lachy Groom, a prominent investor-operator whose influence in Silicon Valley has been on a steep ascent, Physical Intelligence had already secured over $1 billion in funding. Whispers of yet another $1 billion funding round at an $11 billion valuation earlier this spring underscored its considerable market traction. Furthermore, its proprietary AI model, often referred to as a "robot brain," is reportedly a widely adopted tool in robotics research, suggesting a sophisticated level of AI development capable of complex task execution and learning. This strong foundation made the prospect of its acquisition by a leading AI firm not just plausible, but strategically compelling.
While the initial rumor of an outright acquisition may have been imprecise, The Information revealed that Anthropic and Physical Intelligence did, in fact, engage in acquisition talks this past spring. This revelation lends credence to tech blogger Robert Scoble, whose widely disseminated post on X initiated the weekend’s speculation, indicating that while the specifics might have been misconstrued, the underlying event of serious discussions had indeed occurred. The measured response from Physical Intelligence’s CEO, Karol Hausman, who reportedly communicated the inaccuracy of the reports to employees via a Slack message featuring a GIF of a character from "The Office" shaking her head, suggests a nuanced situation rather than a definitive rejection. Lachy Groom, a key figure in Physical Intelligence, has remained notably silent, declining to comment on the matter when approached by TechCrunch.
This strategic maneuvering occurs against the backdrop of significant corporate ambitions. Anthropic has been judicious in its acquisition strategy, completing four known acquisitions this year. OpenAI, however, has been far more aggressive, reportedly acquiring at least 17 companies since 2023. Both AI powerhouses are also on the cusp of major financial milestones, with Anthropic confidentially filing for an IPO on June 1st, followed by OpenAI a week later. These anticipated public offerings are poised to be among the most significant stock debuts in U.S. history, adding a layer of urgency and strategic importance to every move made in the competitive AI arena.
The burgeoning interest in robotics by leading AI companies like Anthropic and OpenAI is not arbitrary; it signifies a pivotal shift in the evolution of artificial intelligence. The prevailing theory is that true superintelligence may necessitate a profound understanding and interaction with the physical world, a domain that cannot be fully grasped through text-based data alone. This belief is echoed in OpenAI’s own history. The company previously experimented with robotics, developing a robotic hand capable of solving a Rubik’s Cube. However, this endeavor was shelved in 2021, with co-founder Wojciech Zaremba later acknowledging that the approach lacked crucial elements for achieving genuine superintelligence. The robotics team was quietly revived in 2024, establishing a humanoid robotics lab in San Francisco. CEO Sam Altman publicly confirmed this resurgence in late May, announcing the launch of "OpenAI Robotics" and outlining a near-term focus on robots for infrastructure work, with a long-term vision of a personal robot for every individual.
Anthropic’s approach to physical world integration, while less hardware-centric than OpenAI’s, has been focused on research and safety. Its internal groups have published a series of research papers that stress-test the frontier capabilities of its AI models, particularly Claude, for safety purposes. A notable experiment, Project Fetch, conducted last November, assessed Claude’s ability to assist non-experts in programming a robot dog. A subsequent phase in June demonstrated remarkable progress, with Anthropic reporting that a newer model completed the same tasks approximately 20 times faster than the best human-plus-Claude team from the previous year. Acquiring a team with established robotics expertise, such as Physical Intelligence, would allow Anthropic to bypass years of foundational research and development.
However, the potential acquisition of Physical Intelligence by Anthropic is complicated by a web of existing investments and relationships. Founded roughly two years ago by Lachy Groom, former Google researchers, and esteemed professors from Stanford and Berkeley, Physical Intelligence’s early investor base mirrors that of OpenAI. Prominent investors include Khosla Ventures and Thrive Capital, both significant backers of OpenAI. Furthermore, Founders Fund, another major investor in OpenAI, was reportedly involved in Physical Intelligence’s most recent funding round earlier this year.
The most significant complication, however, is OpenAI’s direct investment in Physical Intelligence. This makes OpenAI not merely an observer but an active stakeholder in a company that its primary rival, Anthropic, was reportedly in deep acquisition discussions with. This situation raises critical questions about the nature of OpenAI’s early investment. Did it include information rights or a right of first refusal, provisions often negotiated by strategic investors to safeguard against competitors acquiring key assets? Such protective clauses are designed precisely for scenarios like this, where a company’s strategic importance to one player makes it a target for another.
Given these intricate financial and strategic entanglements, the possibility remains that if Physical Intelligence is indeed available for acquisition, OpenAI might possess a more compelling claim. As a shareholder, a close associate of Groom, and a company actively seeking to gain a decisive edge in the robotics domain, OpenAI is strategically positioned to potentially acquire Physical Intelligence itself. When approached for comment regarding these complex dynamics, OpenAI did not respond. The outcome of these behind-the-scenes negotiations will undoubtedly have profound implications for the future trajectory of AI development and the competitive landscape as these tech giants prepare for their public market debuts. The race to imbue AI with physical intelligence is no longer a theoretical pursuit; it is a high-stakes battleground where strategic acquisitions and deep-seated investments are shaping the very definition of artificial general intelligence and its eventual impact on society.
