SoftBank is assembling Japan's industrial elite to build its own foundational artificial intelligence capabilities. Key players including NEC, Honda, Sony, three major banks, Nippon Steel, and Kobe Steel have already invested in a new SoftBank division. Their ambitious goal by the end of the decade is to develop a colossal foundational model with one trillion parameters, focusing specifically on "Physical AI." This refers to AI capable of autonomously controlling robots and machinery.
The rationale behind this mobilization is clear. Even Japan's industrial giants appear weary of relying on foreign AI developments from companies like OpenAI, Anthropic, or Alibaba. As artificial intelligence begins to process increasingly sensitive data, such as the operational status of industrial facilities, concerns about information leakage to overseas servers are mounting. The proposed solution is to maintain control: all data processing will occur on Japanese territory, supported by SoftBank's new data center near Osaka.
The project aims for substantial government funding. Through the NEDO agency, approximately one trillion yen ($6.7 billion) is slated for national AI development over the next five years. It is highly probable that SoftBank's new division will be a primary recipient of these funds.
This Japanese move signals a growing trend toward AI sovereignty, where national interests and data security are prioritized. For global AI leaders, this means intensified competition not only on a technological level but also on the geopolitical stage. The push for domestic AI capabilities, driven by concerns over data control and strategic independence, represents a significant shift in the global AI landscape, forcing established players to re-evaluate their market strategies in the face of resurgent nationalistic technological ambitions.