Japan approves AI robotics strategy targeting a 30% global share and 20 trillion yen market by 2040

Japan's government on March 26, 2026 endorsed a national AI robotics strategy that aims to capture more than 30% of the global market for multi-purpose robots, which it expects to grow to about 60 trillion yen by 2040, giving Japanese companies a market of roughly 20 trillion yen. The document, titled AI Robotics Strategy: Accelerating Social Implementation and Opening Up a Huge Market, was approved without objection at the second meeting of the inter-ministerial liaison council on AI robotics, chaired by Wataru Sakata, Assistant Chief Cabinet Secretary for domestic affairs, according to the council's published minutes.

The strategy's most concrete technical commitment is a domestically developed robot foundation model. The text says the AI Robot Association (AIRoA) will release a beta version as open source around June 2027. The plan names 16 fields for deployment: small-batch manufacturing, shipbuilding, logistics, civil engineering, building construction, infrastructure maintenance, retail, hospitality, elder care, security services, agriculture, forestry, waste processing, disaster response, police work and defense. It also singles out eight common tasks for near-term rollout across those fields, namely indoor and outdoor inspection, indoor and outdoor transport, cleaning, loading and palletizing, handling, and welding and painting.

The measures are grouped in three blocks. On the supply side, the government will support robot makers and system integrators, help startups move from prototypes to volume production in partnership with automotive and industrial supply chains, and strengthen design and manufacturing of motors, reducers and sensors, components the strategy describes as among the best in the world. It calls for a gradual shift to what it terms the Software Defined Robot, in which functions are separated from hardware and defined and updated in software, and for edge semiconductors designed backward from the functions robots need. On the demand side, it links field deployment to a data loop in which installations generate data, data improves models, and better models lower costs and widen adoption. A cross-cutting block creates a Center of Excellence that combines physical test beds with computing resources and simulation, together with a safety certification regime backed by rigorous verification.

Public purchasing is meant to provide the first large orders. The strategy proposes trials and model projects with fire departments for disaster response, a review of design and construction standards in public works based on robot evaluations, and in defense a significant expansion of research and development, trial operation by the Self-Defense Forces and consideration of larger-scale procurement. For the safety and standards work it names the Japan Robot Association (JARA), the Robot Revolution and Industrial IoT Initiative (RRI) and the Japan AI Safety Institute (AISI) as partners.

The council first met on January 16, 2026 with a mandate to finish a strategy within fiscal 2025, which ended on March 31. The March 26 minutes record that a revised roadmap with concrete targets and measures is to follow in May, that the strategy is to be reflected in the government's Basic Policy on economic and fiscal management in June, and that it will be reported to the AI and semiconductor working group of the Japan Growth Strategy Council. Hideaki Ibuki, director general of the Manufacturing Industries Bureau at the Ministry of Economy, Trade and Industry (METI), serves as vice chair.

The document is unusually blunt about where Japan stands. It says Japanese makers held about 70% of the industrial robot market centred on automobiles and semiconductors in 2022, yet trail in AI development aimed at robot deployment and in designing AI and hardware as one system, and that their presence in the expanding service robot market is shrinking. It also concedes that the 2015 New Robot Strategy and the 2019 robot social transformation plan, which funded research, robot-friendly environments and integrator training, produced only limited real-world deployment because they did not create sustained demand.

Industry data released two months earlier show why the government is focused on demand at home. JARA reported on January 23, 2026 that its members' robot orders rose 27.8% in calendar 2025 to 925.8 billion yen and production rose 21.0% to 837.3 billion yen. Exports climbed 36.2% to 694.5 billion yen, while domestic shipments fell 8.9% to 176.3 billion yen. Japanese makers are selling more robots abroad and fewer at home, the opposite of what a labour-short economy would be expected to need.

The international comparison is equally stark. According to the International Federation of Robotics (IFR), China installed 295,000 industrial robots in 2024, more than six times Japan's 44,500, and Chinese suppliers held 57% of their home market. Installations fell 4% in Japan, 9% in the United States and 3% in South Korea that year. Seoul's Ministry of Trade, Industry and Energy launched a K-Humanoid Alliance on April 10, 2025 with the stated aim of making South Korea the leading humanoid robot nation by 2030.

The foundation model at the centre of the plan is a large AI model trained on robot motion, camera and sensor data, so that one model can be adapted to many robot bodies and tasks instead of each application being programmed by hand. AIRoA, the body charged with building it, was founded in December 2024 and launched publicly in March 2025 with 11 regular and five supporting members, including Toyota, NEC, Fujitsu and KDDI, under chair Tetsuya Ogata. The strategy's phrase for the shift it expects is a move in AI competition from scale, meaning web text and raw compute, to integration and operational capability, an area where it argues Japanese factories and component makers have an edge.

For established makers such as FANUC, Yaskawa Electric and Kawasaki Heavy Industries, the document points both ways. Its emphasis on motors, reducers, sensors and edge chips plays to Japan's installed industrial base. Its call for hardware-agnostic software, open interfaces and standardized data formats, which it sets against vertically integrated models, invites competition in the layer above the robot arm, where the strategy admits Japan is behind.

The March text leaves important questions open. It sets no budget, contains no unit or installation targets, and assigns the quantitative goals by sector to the May roadmap. The 16 fields span ministries from health and labour to defense, which makes coordination the main execution risk. The June 2027 date for a beta foundation model will be judged against open models already published by US and Chinese groups.

The next checkpoints are the May roadmap revision, which the minutes say will set concrete targets and measures, and the June Basic Policy, which will show whether the strategy receives priority in the fiscal 2027 budget request.