Japan’s Approach to AI Regulation in 2025

Following Shigeru Ishiba’s appointment as Prime Minister on October 1, 2024, Japan has been charting a cautious but innovation-friendly AI regulation strategy, culminating in a public announcement by the Prime Minister himself on March 3, 2025, expressly favoring a regulatory framework that “promotes innovation” and also “addresses risks.” This is in contrast to some indications in early 2024 where Japan appeared poised to introduce comprehensive, AI-specific regulations in step with moves by the EU and the United States. Consistent with this strategy, Japan’s legislative and regulatory bodies have sought to rely on technologically neutral, sector-specific laws and voluntary industry guidelines, rather than sweeping and restrictive statutes. Their approach seeks to balance fostering innovation with mitigating risks, avoiding excessive or new regulations that could stifle AI development, while addressing societal/security concerns by regulating AI under existing legal frameworks (for privacy, safety, etc.).

• Regulatory Framework and Legislative Measures
  
  In late February 2025, Japan’s Cabinet approved a new AI bill (titled “The Act on the Promotion of Research, Development, and Utilization of Artificial Intelligence-Related Technologies”) that aims at supporting AI research and deployment, and strengthening risk management. Notably, the bill avoids imposing strict new rules or penalties on AI developers. While it generally obligates companies to cooperate with government oversight and to correct any AI-driven infringements on people’s rights, it lacks any material penalty provisions for failure to comply. To coordinate Japan’s AI strategy, the legislation also establishes a high-level AI strategy headquarters led by the Prime Minister, uniting all Cabinet ministers to oversee AI policy implementation and risk response. This new body is expected to monitor emerging AI risks and exercise the authority to request cooperation from relevant agencies and private companies, while continuing to emphasize a supportive regulatory environment overall.

• Cabinet Office Interim Report (February 4, 2025)
  
  Japan’s new course on AI regulation was crystallized in an interim report released on February 4, 2025, by the Cabinet Office’s AI Strategy Council and its AI Policy Study Group, which are tasked with proposing key recommendations for Japan’s AI regulatory framework. By emphasizing the balance between innovation and risk, the report clarifies that AI policies must promote innovation while addressing risks and acknowledges that over-regulation could impede AI’s benefits. Accordingly, the report states that Japan’s AI policy should rely on existing laws, sector-specific regulations, and “soft law” guidelines as much as possible and only introduce new regulations where truly necessary (i.e., where voluntary measures by businesses are insufficient).
  
  International cooperation and AI safety are prominently featured in the Cabinet Office’s recommendations. The report notes that global discussions (through the G7 Hiroshima AI Process, OECD, etc.) are actively shaping AI regulation and urges Japan to align domestic AI measures with internationally accepted principles and frameworks, both to facilitate Japanese companies expanding abroad and to allow Japanese citizens access to AI services worldwide.
  
  On AI safety, the report suggests that the government should establish guidelines (drawing on other countries’ practices and rules) to encourage responsible AI development. This includes voluntarily compliance by businesses with best practices, surveys to understand how companies are managing AI risks, and supporting companies (or intervening under existing laws) based on those findings. It also contemplates (1) the government’s introduction of AI safety evaluations or third-party certification systems to bolster public trust, and (2) in case of serious AI-caused accidents, the government’s collecting of information, alerting the public and developers, and taking steps to prevent recurrences. All of these measures hinge on cooperation with private companies. Thus, legal mechanisms to mandate information sharing and incident reporting from AI operators (both domestic and foreign) are suggested in the report as necessary backstops, but are not strictly mandated.

• Financial Services Agency (FSA) AI Discussion Paper (March 2025)
  
  In the financial sector, Japanese regulators are similarly seeking a “balanced approach” to regulating AI. In March 2025, the FSA published a discussion paper (titled “Preliminary Discussion Points for Promoting the Sound Utilization of AI in the Financial Sector”) that calls on banks and fintech firms to address the potential of AI responsibly and proactively. The FSA highlights that AI offers significant opportunities for the sector, ranging from automating customer service chatbots to streamlining back-office operations and providing data-driven insights for decision making. The FSA also urges caution (especially regarding generative AI) and flags circumstances where AI could be misused: financial fraud, money laundering, or the spread of misinformation in the financial markets. Per these concerns, the FSA hinted that some regulatory measures may be advisable, including developing sector-specific standards for AI model validation, transparency requirements for AI-driven decisions, and oversight to ensure companies’ compliance with existing laws (e.g., on consumer protection and cybersecurity). The FSA paper calls for a “balanced approach” that neither hinders innovation nor ignores risk, and shows the FSA’s willingness to listen and adapt by seeking input on topics like best practices for AI governance, risk management, and areas where further discussion between the agency and companies is needed.

Japan is charting a deliberately moderate course on AI regulation, by prioritizing innovation and international alignment while using existing legal frameworks to manage emerging risks. This approach reflects its administration’s strategic intent to foster AI growth without compromising public trust, safety, or regulatory accountability.