Anthropic Rumored to Acquire Robotics Company Physical Intelligence: AI Giant Enters Physical World

Physical Intelligence was founded in 2023, focusing on developing AI systems that can understand and manipulate the physical world. The company's core technology extends large language model capabilities to the robotics field, enabling robots to understand natural language instructions and execute complex tasks in real environments. Unlike traditional industrial robots, Physical Intelligence's robots have stronger adaptability and learning capabilities, able to autonomously navigate and operate in unseen environments. This 'general robot intelligence' is exactly the capability Anthropic values.

Anthropic's interest in the physical world is not sudden. Last November, Anthropic's internal safety testing team conducted 'Project Fetch,' testing the extent to which Claude models could help non-experts program robot dogs. The second phase in June showed that newer models completed the same tasks roughly 20 times faster than the best human-plus-Claude team from the year before. These studies show that Anthropic's models demonstrate enormous potential in robot control. Acquiring Physical Intelligence would give Anthropic access to professional robotics hardware and engineering teams, accelerating the implementation of its embodied AI strategy.

This rumored acquisition occurs against the backdrop of an AI industry M&A wave. Since 2026, both Anthropic and OpenAI have conducted large-scale acquisitions. Anthropic acquired multiple developer tool companies and AI service providers, while OpenAI focused on acquiring product testing startups. The purpose of these acquisitions is to convert model capabilities into enterprise revenue and expand market influence faster. However, acquiring a robotics company like Physical Intelligence is a different signal—it shows that AI companies are beginning to recognize that the real AI revolution is not just in the digital world, but also in the physical world.

If Anthropic really acquires Physical Intelligence, it will face multiple challenges. First is the difficulty of technical integration—seamlessly combining language models with robot control systems is an extremely complex engineering problem. Second is safety considerations—Anthropic is known for AI safety, and physical world AI systems involve more direct safety risks, such as robot operation errors that could cause personal injury. Third is market competition—tech giants like Google and NVIDIA are also actively deploying in the embodied AI field, and Anthropic needs to find differentiated competitive advantages. Nevertheless, if successful, this acquisition would make Anthropic the first AI company to simultaneously possess top-tier language models and robotics technology.