Japan’s AI Basic Plan to Promote AI Use in Public Institutions

Japan sets out first national AI basic plan: public sector to lead adoption

Japan has drafted its first national AI Basic Plan, a roadmap the government aims to finalize within the year to make the country the most AI friendly place for development and use. The draft calls for public institutions to lead by example. Ministries, agencies, and municipalities are expected to adopt artificial intelligence to improve efficiency and service quality, with defense applications also on the table. The plan seeks to balance rapid innovation with responsible risk management at a time when usage in Japan remains modest. Surveys indicate that about 20 percent of individuals and roughly half of companies used AI last year, and only 26.7 percent of people reported using generative AI in fiscal 2024, far below rates in the United States and China. The draft is part of a new legal and policy framework that includes an AI law enacted in May 2025 and a Prime Minister led AI Strategy Headquarters.

The document sets four basic policies: accelerate AI utilization, strengthen development capacity, lead in AI governance, and keep innovation moving toward an AI integrated society. It stresses human AI collaboration, a strong data foundation, and clear rules for issues like copyright and civil liability. It also flags risks, including erroneous outputs, disinformation, privacy harms, and national security challenges. The draft will be presented to the AI Strategy Headquarters chaired by Prime Minister Shigeru Ishiba. An expert panel will then refine it, with Cabinet approval targeted for later this year. Authorities intend to investigate serious rights infringements linked to AI use and help shape international rules, including through the Hiroshima Principles for safe and trustworthy AI.

What the draft says: four pillars and why they matter

The plan’s four pillars knit together technology, policy, and institutions. First, government bodies should deploy AI to streamline work and improve public services. Second, development capacity must be strengthened by leveraging high quality data, compute resources, and talent. Third, Japan intends to take a leadership role in AI governance by aligning domestic guidelines with international principles and by equipping authorities to investigate serious incidents. Fourth, society should support continuous innovation and human AI collaboration through education, workforce training, and institutional reform.

Accelerating AI use across government

The draft instructs national and local agencies to lead the shift to AI enhanced public services. Practical targets include faster document processing, automated translation and summarization, case triage for benefits and permits, and improved call center support through AI assistants that escalate complex issues to human staff. Municipalities can use AI for disaster preparedness, hazard mapping, evacuation planning, and real time information during storms and earthquakes. Health systems can prioritize screenings, flag potential adverse events, and optimize scheduling. Transport authorities can improve traffic signal timing and fleet dispatch. Courts and regulators can use tools to search case law and detect patterns in filings. These applications aim to reduce backlogs, cut routine workload, and deliver quicker responses to citizens.

Japan faces labor shortages and an aging population. Public institutions will need tools that let small teams handle more work with consistent quality. AI can test different policy scenarios quickly and scan large data sets for fraud, tax gaps, or safety violations. The draft frames these uses as collaboration, not substitution. Staff remain accountable for decisions and oversight, while AI handles information heavy sub tasks, such as spotting anomalies or suggesting first drafts that human experts review.

Building development capacity

The draft calls high quality data one of Japan’s strengths. Industries like manufacturing, health, and transport produce rich, structured information that can raise AI accuracy if curated well. The plan highlights the need for robust data pipelines, consistent labeling, and privacy safeguards. It encourages more open data where feasible, so that academia, startups, and established firms can build and evaluate models. Investment in compute, storage, and networking will also matter, along with support for models tuned to Japanese language and context.

Talent and tools are core to capacity. The draft points to education and reskilling programs so civil servants, engineers, and domain specialists can use AI responsibly. It also implies procurement reforms that give agencies flexibility to run pilots, evaluate performance, and set clear requirements for security, privacy, and quality. A broader supplier base, including small firms and research consortia, can accelerate experimentation while maintaining safeguards.

Leading in AI governance

Japan’s AI law, enacted in May 2025, sets the framework for a national strategy and creates institutions to coordinate it. The law establishes an AI Strategy Headquarters in the Cabinet and an AI Strategy Center to guide the Fundamental AI Plan, develop guidelines, and coordinate across ministries. It places responsibilities on AI developers, providers, and business users to make reasonable efforts to follow government principles and cooperate with investigations. The approach relies on guidance and transparency rather than punitive fines, while reserving authority for government inquiries when serious risks to rights or safety arise.

Key rules remain anchored in existing law. The Act on the Protection of Personal Information governs consent, data minimization, breach reporting, and cross border transfers. Copyright questions tied to training data and outputs are under review, with the draft emphasizing the need to protect property while enabling fair use and innovation. Japan also aims to curb disinformation and harmful content, including with faster takedown mechanisms under recent legislation for information distribution providers. The draft aligns with international efforts, including the Hiroshima International Guiding Principles, to promote safe, trustworthy, and human centric AI.

Chief Cabinet Secretary Yoshimasa Hayashi has described the government’s stance as a balance between opportunity and risk. He pointed to productivity gains and labor relief alongside threats from misuse and criminal activity.

“AI offers benefits like improved productivity and solutions to labor shortages, it also poses risks such as disinformation and sophisticated criminal activity.”

The governance pillar signals increased transparency and active oversight. Agencies will survey AI use, publish guidance, and, in serious cases, investigate incidents and disclose relevant information. Businesses and research institutions will be expected to document AI life cycles, manage data rights, and address bias, safety, and security in design and deployment.

Continuous innovation toward an AI society

The fourth pillar emphasizes human AI collaboration and ongoing institutional change. That includes reskilling programs for public sector staff and incentives for partnerships between universities, startups, and established companies. Public services can benefit from the convergence of AI with robotics and sensors, often described as physical AI. In areas such as eldercare, logistics, and inspection, AI enabled systems that perceive and act in the physical world can shift routine tasks from scarce personnel to machines supervised by humans.

The draft also acknowledges that laws, benefits systems, and workforce policies will need regular updates to keep pace with rapid advances. Public communication and transparency will be vital to build trust. Government will need to explain where and why AI is used, what benefits and risks are expected, and how citizens can seek redress if something goes wrong.

Why Japan is moving now

Japan’s lag in AI adoption is a core concern. Surveys show roughly one in five individuals used AI last year, and only about half of companies reported use. For generative AI specifically, 26.7 percent of individuals in fiscal 2024 reported use, compared with much higher levels in the United States and China. Private AI investment in Japan also trails other major economies by a wide margin. Policymakers see a risk that domestic firms could fall behind competitors that have integrated AI into products, operations, and customer service.

Several structural factors compound the gap. Many organizations retain paper heavy processes and legacy systems. Data sets sit in silos and in formats that are hard to integrate. Procurement cycles favor large, multi year projects over small pilots that scale on merit. Compliance risk is often managed through avoidance rather than mitigation. Language and domain nuance can also limit the transfer of foreign models without careful tuning. The Basic Plan aims to remove these barriers through clear guidance, common tools, and institutional support so agencies and firms can adopt AI responsibly and at speed.

Risk management and law: from soft guidance to a national framework

The AI law defines policy goals, institutions, and responsibilities for the research, development, and use of AI related technologies. It covers developers, providers, and business users, including foreign operators that supply AI used in Japan. The law’s core principles stress productivity, transparency, privacy, fairness, and safety. It empowers the government to survey practices, issue guidance, and investigate serious cases involving the infringement of rights and interests. While it does not impose broad penalties similar to those in the European Union, it gives authorities a clear path to request information, provide instructions, and disclose findings in significant cases.

The Basic Plan builds on that structure. It targets clear civil liability frameworks for AI use and stronger protection for intellectual property, especially in copyright sensitive fields like media, software, and design. It encourages lifecycle risk management, from data collection and model training to deployment and monitoring. Privacy compliance under the APPI is central. Agencies and vendors will need to justify data use, apply techniques like pseudonymization, and maintain robust access controls. Transparency to users and the public will be part of accountability, especially when automated systems affect benefits, healthcare, or public safety.

Defense and national security dimensions

The draft proposes adopting AI to strengthen defense capabilities and protect critical infrastructure. Potential uses include logistics planning, maintenance prediction, cyber defense, analysis of satellite and sensor feeds, and decision support for complex operations. These tools can expand the capacity of defense personnel and improve reaction speed. The plan also warns that AI can be misused, from deepfake propaganda to advanced cyberattacks and autonomous threats. A national framework that pairs innovation with security controls is presented as a strategic necessity.

At the headquarters meeting on AI strategy, Prime Minister Shigeru Ishiba underscored the security stakes and the need for urgency.

“AI is very crucial in terms of security,” he said, adding that Japan must “implement support measures promptly” as the global development race intensifies.

Defense adoption will require strict evaluation of reliability, robust testing against adversarial manipulation, and careful human oversight. The plan’s focus on high quality data and disciplined governance is consistent with these needs. It also links to international cooperation on safe AI practices and responsible military use to reduce risk escalation.

How the plan will be built and who will run it

The AI Strategy Headquarters, established in the Cabinet Office and chaired by the Prime Minister, is steering the process. All Cabinet ministers participate. The draft plan will be presented to the headquarters, then an expert panel of researchers, industry leaders, and policy specialists will refine details before the Cabinet seeks approval within the year. The AI Strategy Center will coordinate implementation across ministries and agencies and will continue to publish guidelines and surveys that track practice.

Several bodies will shape rules and practice. The Ministry of Economy, Trade and Industry and the Ministry of Internal Affairs and Communications co author business guidelines. The Agency for Cultural Affairs and the Personal Information Protection Commission guide copyright and privacy. Sectoral regulators, including in finance and health, will align domain specific guidance with the overarching plan. The Digital Agency and local governments will be central to deployment, since many services touch residents directly.

What public institutions can do first

Agencies do not need to wait for final approval to prepare. The draft signals priorities that teams can begin to act on now. Leaders can identify high value use cases where AI can save time or improve outcomes, and where quality data and clear oversight are available. They can build governance into projects from the start and plan for audits, documentation, and user communication. They can also invest in skills so staff can evaluate AI outputs and understand constraints.

• Assign a senior official to coordinate AI across the agency and to align projects with privacy, security, and accessibility requirements.
• Map data assets and classify sensitivity. Establish processes for data quality, labeling, and access control consistent with the APPI.
• Start with narrow pilots that have measurable goals, like reducing document turnaround time or improving contact center resolution, and scale only after evaluation.
• Include clear contract clauses with vendors on data rights, copyright, liability, incident reporting, and model quality assurance.
• Explain to users when and how AI is used, provide human support paths, and record feedback for continuous improvement.

International context: how Japan compares and can influence rules

Major economies are taking different routes to AI oversight. The European Union is moving toward a comprehensive legal regime for high risk systems. The United States relies more on agency guidance and executive action. Japan’s approach sits between them. It combines a national framework with soft law and sectoral guidance, while reserving investigations and public disclosures for serious cases. The draft also puts Japan in a leadership role on international principles, building on the Hiroshima initiative. That position can help shape norms for disinformation response, copyright, safety evaluation, and transparency.

Japan’s strength in quality manufacturing, robotics, and precise data collection gives it a foundation to lead in trustworthy AI. If the Basic Plan succeeds in modernizing procurement, unlocking data, and aligning governance, Japan can boost productivity at home and set practical standards that others adopt.

The Bottom Line

• Japan’s first AI Basic Plan sets four policies, from acceleration of use to leadership in governance and continuous innovation.
• Government institutions and municipalities are expected to lead AI adoption, with defense uses proposed to strengthen capabilities.
• Domestic use of AI remains low, with around 20 percent of individuals and about half of companies using AI, and 26.7 percent reporting use of generative tools in fiscal 2024.
• The AI law enacted in May 2025 created an AI Strategy Headquarters and an AI Strategy Center to design and coordinate the Fundamental AI Plan.
• The draft targets clear civil liability and stronger copyright protection while keeping privacy compliance under the APPI at the core.
• Authorities plan to investigate serious rights infringements involving AI and to disclose information in significant cases, while prioritizing guidance over punitive fines.
• High quality data is cited as a national advantage that can improve model accuracy and reliability.
• The draft will be refined by an expert panel and is slated for Cabinet approval within the year.