Japan - AI Research and Development Act (2025)

This Act serves as Japan's foundational national law for Artificial Intelligence. Its primary goal is to foster innovation in AI research, development, and practical application, while simultaneously addressing potential societal and safety risks. Unlike some other regulations, it adopts a promotional and coordinative approach, focusing on government leadership, guidelines, and cooperation rather than imposing extensive new prohibitions or criminal penalties.

The Act broadly defines its scope to include "artificial intelligence‑related technologies." It applies to any individual or entity involved in researching, developing, or utilizing these technologies in connection with Japan, explicitly including foreign entities whose activities impact Japanese businesses or citizens. Further technical definitions and operational meanings are expected to be clarified through implementing guidelines and the forthcoming AI Basic Plan.

This provision establishes the Artificial Intelligence Strategic Headquarters, a high-level governmental body central to Japan's AI policy. Chaired by the Prime Minister and comprising all Cabinet ministers, this Headquarters is responsible for coordinating and implementing national AI strategy.

A key function of the Headquarters is to prepare and periodically update the 'AI Basic Plan.' This plan will articulate the government's objectives for AI, identify priority sectors for development, outline measures for research and infrastructure investment, and define strategies for international cooperation. All ministries and agencies are mandated to cooperate with the Headquarters and to develop their own sector- or function-specific guidelines in alignment with the overarching AI Basic Plan.

The Act outlines several key areas where the government will focus its efforts to promote AI. These include boosting research and development, expanding and sharing essential infrastructure like data centers, fostering human resource development and AI education, and creating guidelines to ensure AI systems are transparent, fair, and safe. It also emphasizes monitoring risks, investigating incidents, and actively participating in international discussions on AI norms.

Crucially, this provision defines the roles of various stakeholders. National and local governments are expected to implement enabling measures. Research institutions are tasked with strengthening fundamental research and collaboration. Business operators are encouraged to cooperate with government initiatives and voluntarily comply with published guidelines. Finally, citizens are recognized as stakeholders with a right to safety and information regarding AI.

This article details the primary mechanisms through which the AI Act will be implemented. These include the AI Strategic Headquarters and the AI Basic Plan, which provide overall policy direction and set priorities. Secondly, various ministries and agencies will issue specific guidelines to operationalize responsibilities and define best practices for different sectors or functions. Thirdly, the government will establish investigative and monitoring functions to gather information on incidents that could affect human rights, safety, or public interests, and to provide guidance based on its findings.

The Act empowers the government to publish guidance, offer technical support, promote investment in AI infrastructure, and solicit voluntary cooperation from stakeholders. It explicitly states that the Act does not rely on new criminal sanctions but rather on these supportive and cooperative tools to achieve its objectives.

The Act mandates a continuous process of monitoring and evaluation to ensure AI technologies are developed and utilized responsibly. This includes the systematic collection of information and the investigation of incidents where AI systems may have negatively impacted human rights or broader public interests. Based on these investigations, the government is authorized to issue public reports or provide specific guidance to operators.

This monitoring framework is designed to be adaptive, meaning ministries will continuously track technological advancements and market practices. The AI Basic Plan itself will be periodically updated to reflect these developments and evolving societal needs. The Act also references international reporting frameworks, such as those from the Hiroshima AI Process, as part of its commitment to transparency and international accountability.

A key characteristic of Japan's AI Act is its non-punitive enforcement philosophy. Unlike many other regulatory frameworks, this Act does not introduce a new system of fines or criminal penalties specifically for AI-related activities. Instead, enforcement primarily relies on mechanisms such as issuing guidance, making public announcements, conducting investigations, and requesting cooperation from relevant parties.

It is crucial to understand that while the AI Act itself avoids new penalties, it does not override or displace existing legal frameworks. This means that sectoral laws, civil liability (e.g., tort or contract law), criminal statutes, and data protection regulations (like the Act on the Protection of Personal Information) continue to apply where relevant to AI activities. Any harms or non-compliance related to AI could still lead to liability under these pre-existing laws.

This provision clarifies that the AI Act is designed to work in conjunction with, rather than replace, Japan's existing legal landscape. It explicitly complements a range of established laws, including those governing data protection (such as the Act on the Protection of Personal Information, or APPI), consumer protection, product safety, telecommunications, network security, intellectual property, and unfair competition, as well as the Penal Code.

The Act specifically directs various ministries to coordinate these existing legal instruments. Their role will be to produce guidelines that interpret how these diverse sectoral laws apply to AI systems. This approach positions the AI Act as a foundational framework that enables targeted, sector-specific regulation as needed, ensuring a cohesive and comprehensive legal environment for AI.

Japan's AI Act is framed with a strong emphasis on international alignment and cooperation. The government explicitly positions this Act as consistent with global processes, particularly highlighting its connection to the Hiroshima AI Process, which emerged from the 2023 G7 Summit, as well as standards from the OECD and other international bodies.

This provision specifically tasks the Japanese government with actively participating in international norm-setting efforts for AI. Furthermore, it mandates that the guidelines formulated under this Act must be in accordance with international norms. This reflects Japan's commitment to being a constructive partner in shaping global AI governance and ensuring its domestic framework is harmonized with evolving international standards.