Japan

The law firm Clifford Chance analyzes a bill approved by Japan's Cabinet on 28 February 2025, a proposed statute on promoting AI-related research, development and use. The firm describes the bill as unexpectedly short and abstract, aimed at encouraging AI work and ethical use while protecting citizens' rights, relying on cooperation between the state, companies and the public instead of strict requirements.

Most provisions set general principles for national and local governments and businesses and describe government duties in broad terms. Businesses and citizens are expected to cooperate on a reasonable-efforts basis but face no specific hard obligations. The firm contrasts this soft, high-level approach with the detailed EU AI Act and suggests it may answer industry worries about stifling innovation. The bill sets out no direct penalties.

Article 16 requires the government to gather information on AI trends at home and abroad, analyze cases where citizens' rights were infringed by wrongful or improper AI use, consider countermeasures, and then take necessary measures such as guidance, advice and information. Reports, including from the Nikkei, say the government will be able to publicly name developers and users that significantly infringe rights, which Clifford Chance sees as a reputational risk. The bill also provides for an AI Strategy Headquarters led by the Prime Minister with all ministers as members, and an AI Basic Plan, and addresses research facilities, human resources, education and international norms.

The House of Representatives was debating the bill. If consensus is reached, it could pass before the current Diet session closes on 22 June 2025 (absent an extension), and key provisions would take effect on publication. Criteria for acting against businesses are expected afterwards, possibly as guidelines subject to public consultation.