Japan AI Governance Profile

Executive summary

Japan’s 2025 AI Act establishes a statutory national framework for promoting AI research, development and use and for coordinating government policy. It became fully effective on 1 September 2025 and provides for the AI Strategic Headquarters and AI Basic Plan. The government updated the Basic Plan on 14 July 2026 around the concept of trusted AI and society-wide AI transformation. For enterprises, the AI Guidelines for Business and Japan AISI’s safety-evaluation guidance are operationally important but should be classified as non-binding guidance and evaluation infrastructure unless another legal instrument independently creates an obligation.

Governance architecture

Japan’s model is promotion-led statutory coordination. The AI Act creates national institutions, responsibilities and planning mechanisms, while the AI Basic Plan, business guidance, existing sector and data/IP law, and AISI evaluation infrastructure carry much of the operational governance burden. The statute should not be presented as an EU-style enterprise conformity regime.

Major policies and frameworks

Policy timeline

Enterprise implications

Enterprises should not build compliance around an assumed EU-style risk-classification regime. They should track sectoral, data and intellectual-property law, use official business guidance as an implementation benchmark, and use AISI evaluation guidance for assurance and testing even where it is not legally mandatory. Monitor whether future policy converts guidance into more prescriptive sector requirements.

Observatory interpretation

Japan’s model emphasizes national coordination, trusted adoption and evaluation capability. Institutions need people able to translate non-binding guidance, sector rules and technical evaluation methods into internal governance rather than relying on a single statutory conformity checklist.

Official resources

Update log