Tech Desk 
Sep 19, 2025
The Japanese government plans to build an artificial intelligence (AI) model in Japan using advanced resources and technology to lessen the reliance on AI built in the United States (US) and China.
Reportedly, the aim is to ensure Japan’s data security and to provide AI-generated information that reflects Japan’s culture, customs, and history.
AI models from the US and China are the dominant players in the world market, and services such as OpenAI and Alibaba are the front-runners in generative AI, a form of technology that can produce text, images, and additional content.
However, Japanese officials have expressed concern about the use of foreign-produced models because of the risk of sending sensitive data offshore and the risk of generating inaccurate information on the internet.
To counter these issues, the Japanese government will provide a platform and funding to a consortium of Japanese companies to create learning data, train models, and operate domestic data centres.
Specifically, the National Institute of Information and Communications Technology (NICT) will provide two decades’ worth of Japanese-language data, while Preferred Networks, Inc., which is located and headquartered in Tokyo, is tasked with designing the large language model.
IT firm Sakura Internet Inc. is likely to host the model on servers also located in Japan.
Officials say the new AI model will provide highly reliable answers on Japanese topics — from culture and customs to legal systems and geography — and will be used by central and local governments as well as private companies.
