Overview Of Japan's AI Governance: What Global Companies Need To Know

I.Introduction: The Current State of Japan's AI Governance

In May 2025, Japan's Parliament passed the Act on Promotion of Research and Development, and Utilization of Artificial Intelligence-Related Technologies (the "AI Promotion Act")1 in response to the rapid advancement of AI technologies. This law establishes AI as a foundational technology for the development of Japan's economy and society and sets out the government's basic principles to promote consistent R&D in AI-related technologies— from basic research to practical applications.

The AI Promotion Act is not a regulatory framework that is focused on the private sector; rather, it articulates policy objectives to foster AI innovation in general while addressing its potential risks. A key feature of Japan's AI governance is its reliance on a soft law approach, which avoids legally binding penalties and instead provides guidance on AI governance through government-issued guidelines. This approach encourages voluntary corporate initiatives while avoiding excessive regulations that could hinder innovation.

In addition, many AI-related risks are already being addressed under existing legal frameworks. For example, the following laws apply to AI use:

• Act on the Protection of Personal Information (Act No. 57 of 2003, as amended): Regulates the collection, analysis and use of personal data by AI.
• Copyright Act (Act No. 48 of 1970, as amended): Governs the use of copyrighted works in the training and output of generative AI, with certain exceptions for data analysis and machine learning.
• Consumer protection laws: Regulate misleading representations and unfair solicitations.

Thus, Japan's AI governance operates wholistically by combining new measures—such as the AI Promotion Act and governmental guidelines—with the application of existing laws.

Japan's approach differs significantly from the EU's binding regulatory framework, as exemplified by the AI Act, which classifies AI systems by risk level and imposes strict legal obligations and penalties on high-risk AI. It also diverges from the U.S. model, which emphasizes a voluntary framework and flexible guidelines to foster innovation (e.g., the AI Risk Management Framework of the National Institute of Standards and Technology and the Blueprint for an AI Bill of Rights).

Japan has adopted a hybrid model that combines soft-law instruments—such as government-issued AI guidelines— with the application of existing sector-specific legislation. This approach aims to balance the promotion of innovation with the mitigation of risks and the building of public trust. Through this model, Japan seeks to play an active role in shaping international discussions on AI governance. This article provides an overview of Japan's AI governance by focusing on three key topics:

• the AI Promotion Act
• the AI Business Guidelines
• the application of existing laws to AI

II. Overview of the AI Promotion Act: Legal Framework and Policy Significance

The AI Promotion Act is a framework law that outlines Japan's basic policy framework for AI. Instead of imposing specific regulations and penalties, the Act sets out a national policy to comprehensively and systematically promote the research, development and utilization of AI-related technologies.

1. Purpose and Basic Principles

Article 1 (Purpose) of the AI Promotion Act defines the law's objective as contributing to the improvement of people's lives and sound development of the national economy, recognizing AI as a foundational technology for Japan's economic and social progress. Article 3 (Basic Principles) of the AI Promotion Act outlines the following:

• Importance for economic society and national security—keeping R&D capabilities and enhancing international competitiveness
• Comprehensive and systematic promotion from basic research to utilization
• Ensuring transparency, etc., for appropriate R&D and utilization
• A leadership role in international cooperation

2. Structure of Basic Measures (Chapter 2)

The AI Promotion Act organizes the government's responsibilities into a set of policy measures, including:

• Article 11: Promotion of integrated R&D from basic research to practical application
• Article 12: Development and shared use of data centers and related infrastructure
• Article 13: Establishment of guidelines aligned with international standards
• Articles 14–15: Development of advanced AI talent and promotion of education
• Article 16: Investigation and analysis of risk cases, and provision of guidance to businesses
• Article 17: Participation in international rule-making

These measures aim to strengthen Japan's technological foundation while ensuring social acceptance and international alignment.

3. Policy Implementation System: Basic Plan and Strategic Headquarters (Chapters 3-4)

The government must adopt a Basic Plan for Artificial Intelligence in line with the AI Promotion Act's basic principles. The plan will be approved by the Cabinet and made public by the Prime Minister (Article 18).

The Cabinet will also establish the AI Strategy Headquarters (Articles 19–28), chaired by the Prime Minister and composed of relevant ministers. This headquarters will serve as the cross-ministerial command center for AI policy-making, enabling agile and flexible implementation.

4. Duties of Private Businesses and Citizens (Articles 7-8)

The AI Promotion Act assigns roles not only to the government but also to other stakeholders:

• Private Businesses (Article 7): Encouraged to actively utilize AI technologies to improve efficiency, enhance operations and create new industries, while cooperating with measures implemented by the national and local governments
• Citizens (Article 8): Encouraged to deepen their understanding of and interest in AI technologies, and make efforts to cooperate with measures implemented by the national and local governments

5. Legal Nature and Institutional Design

The AI Promotion Act is a non-binding framework law, imposing only "effort-based" and "cooperation" obligations on businesses and citizens. This design avoids excessive regulation that could hinder innovation, while allowing the government to provide direction through a soft law framework. However, failure to cooperate may lead to being subject to guidance, recommendations, or other measures from the authorities. In policy discussions, the possibility of publicly disclosing the names of non-cooperative businesses has also been suggested, but there is no formal mechanism yet for it.

III. Overview of the AI Guidelines for Business: A "Living Document" Bridging Practice and Ethics

While the AI Promotion Act sets the national direction as a framework law, the practical implementation of AI governance is supported by the government-issued AI Guidelines for Business Ver 1.0, published by the Ministry of Internal Affairs and Communications and the Ministry of Economy, Trade and Industry on April 19, 2024.

These guidelines, as a form of soft law without binding legal effect, systematically set out the voluntary measures expected of businesses involved in the development, provision and use of AI. The guidelines aim to present a unified set of guiding principles for AI governance in Japan to promote the safe and secure use of AI, help businesses fully recognize AI-related risks in line with international trends and stakeholders' concerns, and encourage voluntary countermeasures across the entire lifecycle of AI.

1. Human-Centered AI Principles as the Foundation

The core value of the AI Guidelines for Business is rooted in Japan's "Basic Philosophies" as articulated in its human-centered AI framework. These philosophies envision AI as a public good that fosters transformative innovation and global sustainability, anchored in three fundamental values: respect for human dignity, diversity and inclusion, and sustainability.

Based on this principle, the guidelines set out ten common principles: human dignity, inclusion of diversity, sustainability, safety, fairness, privacy protection, transparency, accountability, education and literacy, and innovation. These principles are aligned with Japan's constitutional values as well as international human rights norms and call for ethical and social considerations throughout the AI lifecycle.

2. Responsibilities by Stakeholder Type

A distinctive feature of the AI Guidelines for Business is that they go beyond listing abstract principles and provide concrete considerations tailored to three types of stakeholders: AI developers, providers and users. For example, developers are expected to ensure data accuracy and analyze bias; providers are expected to ensure service transparency and accountability; and users are responsible for appropriate use and establishing educational frameworks.

The guidelines also adopt a risk-based approach, recommending that the level of response be flexibly adjusted according to the purpose and impact of the AI system. This reflects a realistic design philosophy that seeks to secure public trust while avoiding overregulation.

3. Evolution as a "Living Document"

The drafting process of the AI Guidelines for Business involved diverse stakeholders, including educational and research institutions, private companies and civil society organizations, and incorporated public comments and discussions to ensure both practical relevance and legitimacy. The guidelines serve as a "living document" to be reviewed and, where appropriate, revised in response to technological and societal changes, with particular attention to maintaining consistency with international standards.

IV.Application of Existing Laws to AI in Japan

While the newly introduced AI Promotion Act and the AI Business Guidelines provide new policy frameworks, Japan also applies existing legislation to AI-related activities. Although the guidelines themselves are soft law without any binding effect, any use of AI that breaches existing statutes may be deemed unlawful and could trigger civil or criminal liability. Below are examples of legal issues frequently encountered in corporate practice:

1. Act on the Protection of Personal Information ("APPI"): Can personal data be input into AI?

The APPI requires that the purpose of use of personal information be clearly specified and notified or disclosed to the individual. If AI processes personal data in a way that exceeds or deviates from the stated purpose, it may constitute a violation.

Practical Interpretation:

• Explicitly include AI-related processing within the stated purpose of data use.
• When transferring personal data overseas, obtain the individual's consent or ensure that the transfer is made under a legally compliant cross-border data protection framework.

2.Copyright Act: Who owns the copyright in AI-generated outputs?

If AI-generated outputs are substantially similar to existing copyrighted works, then there is a risk of infringement.

Practical Interpretation:

• Infringement is assessed based on two factors: similarity to a pre-existing work and the derivation from (or dependence on) it.
• Where copyrighted works are included in the training data, derivation may be presumed depending on the nature and extent of their use.
• For AI-generated outputs to qualify for copyright protection under Japanese law, there must be human creative involvement—for example, through specific and substantive prompt design or other creative choices that materially influence the output.

3. Consumer Protection Laws: What if AI provides incorrect information to consumers?

Misleading advertisements or product descriptions generated by AI may violate the Act against Unjustifiable Premiums and Misleading Representations (Act No. 134 of 1962, as amended), or, in certain cases, the Consumer Contract Act (Act No. 61 of 2000). Under these laws, a company may be held liable if it publishes or relies on AI-generated output without prior adequate human verification, particularly where the information could mislead consumers or influence their transactional decisions.

Practical Interpretation:

• Establish internal procedures to ensure that all AI generated consumer-facing content is subject to human review and verification before being released.
• For information that could materially influence consumer decisions—such as pricing, product functionality or safety—adopt formal internal rules, maintain records of the review process, and document the basis for factual claims to demonstrate compliance in case challenged by regulators.

V. Conclusion

• Characteristics of Japan's AI Governance

• Japan's AI governance model is characterized by the integration of soft law and existing legislation. The AI Promotion Act provides a foundational direction of government and national policy, while the AI Business Guidelines offer practical behavioral standards to the private sector. This framework respects corporate autonomy while ensuring public trust, with the government supporting enforcement through recommendations, guidance and information dissemination rather than penalties.

  In addition, Japan applies existing laws—such as the APPI, Copyright Act, and consumer protection laws— to AI-related issues. This approach enables flexible and effective responses without creating entirely new AI specific regulations.

  2. Practical Implications for Global Companies

  Japan's approach to AI governance is attracting growing international attention for its attempt to balance innovation with public trust. Given its characteristics, global companies operating in Japan must pay close attention to the local regulatory landscape. While Japan imposes relatively few binding legal obligations, the government-led soft law framework—particularly the AI Business Guidelines—can significantly influence corporate reputation and public perception.

  Moreover, companies must carefully assess how existing Japanese laws apply to their AI-related activities. Practices that are lawful in one jurisdiction may violate Japanese laws, such as the APPI or the Copyright Act, while certain actions that are restricted elsewhere may be permissible in Japan.

  Accordingly, companies operating in Japan are strongly advised to review and update their internal policies on AI use. This should include clearly defining the purposes for using personal data, implementing robust trade secret management and protection systems, and verifying the accuracy and reliability of consumer facing AI-generated outputs. Such measures go beyond mere legal compliance—they are critical to protecting corporate reputation and maintaining public trust.

  The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.