This article introduces 8 top Japanese and U.S. stocks focused on physical AI. It reviews the latest developments, including collaboration announcements in July and a global competition in August.
Since the start of 2026, “physical AI,” in which AI controls robots and machinery, has emerged as a major investment theme. This article examines 8 top Japanese stocks, along with under-the-radar names and the potential for ten-bagger returns. In July, leading Japanese companies Fujitsu, FANUC, Yaskawa Electric and Kawasaki Heavy Industries announced collaborations using NVIDIA technology, raising interest in related stocks in Japan. In August, the World Humanoid Robot Sports Tournament was held in China, and footage of robots being deployed in real-world settings attracted attention.
After explaining what physical AI is and why it is attracting attention, the article organizes the theme around 3 elements: the AI model that serves as the brain, the robot itself that serves as the body, and implementation for real-world applications. The market is expected to expand against a backdrop of labor shortages, with estimates indicating that it could reach approximately ¥60 trillion around 2040. Deployment in real-world settings has already begun in some areas. The development race between the United States and China is accelerating, with CES 2026 identified as a potential turning point and expectations that humanoid robot production in China will expand rapidly.
The “8 top picks” comprise 5 Japanese stocks, including Hitachi, and 3 U.S. stocks, with the article discussing each company’s current position and strengths. The top Japanese picks are FANUC, Yaskawa Electric, Mitsubishi Electric, Hitachi and Fujitsu, while the top global picks are Microsoft, NVIDIA and Tesla. The Japanese government has outlined a plan for combined public- and private-sector investment totaling ¥10.5 trillion, with the aim of accelerating real-world deployment by leveraging Japan’s strengths in components and on-site operations.
The “laggards and under-the-radar names” section covers 2 companies, Nippon Seiko and Kudan, outlining both their strengths in supplying components such as bearings and ball screws and in spatial-recognition software, as well as the investment risks associated with continued losses. On the potential for ten-bagger returns, the article takes the view that a tenfold gain is difficult for large-cap stocks but opportunities may exist among small and mid-cap stocks, and concludes that investments should be limited to a portion of one’s assets.