Japan Sets Strict Guidelines for Responsible AI Use in Defense Equipment

Japan Unveils New AI Guidelines for Defense Equipment Development

On June 6, Japan’s Acquisition, Technology & Logistics Agency (ATLA) introduced the country's first official guidelines for the use of artificial intelligence (AI) in defense equipment research and development (R&D). These guidelines focus on maintaining proper human involvement while managing risks associated with AI-integrated defense systems. They also clarify expectations for businesses working with the Ministry of Defense (JMOD) and the Self-Defense Forces (JSDF), encouraging responsible innovation in this critical sector.

Addressing Legal, Ethical, and Operational Risks

The guidelines respond to growing concerns about autonomous AI systems, especially "Lethal Autonomous Weapons Systems" (LAWS). Japan defines LAWS as systems capable of selecting and engaging targets with lethal force without further human input. Tokyo firmly opposes the development of such fully autonomous lethal weapons and actively promotes their global ban.

Japan’s approach emphasizes a "human-centric principle," ensuring meaningful human control over AI-enabled defense equipment. The guidelines require compliance with domestic and international laws, including international humanitarian law (IHL), reinforcing Japan’s position against autonomous weapons operating without human oversight.

Background and Context

These JMOD guidelines come shortly after the enactment of a broader AI bill on May 28, which aims to boost Japan's AI capabilities while regulating risks across all sectors. The defense-specific guidelines provide transparency and predictability for companies interested in AI R&D related to defense equipment.

Risk Management Process Explained

The guidelines prescribe a three-step risk management process for AI-powered defense projects:

• 1. Classification: Defense equipment is categorized as either "high risk" or "low risk" based on how much AI influences its destructive capability. High-risk projects face stricter review protocols.
• 2. Legal and Policy Review: A dedicated board evaluates projects against two key criteria:
  • A-1: Compliance with international and domestic laws, including IHL.
  • A-2: Ensuring the system is not a fully autonomous lethal weapon operating without meaningful human judgment and control.
• 3. Technical Review: For approved projects, a Technical Review Board assesses seven technical requirements (B-1 to B-7) ranging from human responsibility and operator understanding to system transparency, bias mitigation, reliability, safety, and legal compliance.

Seven Technical Requirements for AI Systems

• B-1: Clear human responsibility with operator control mechanisms.
• B-2: Operator understanding supported by safeguards against misuse or AI malfunction.
• B-3: Measures to identify and reduce bias in AI decision-making.
• B-4: Full documentation of AI design, algorithms, and training data for transparency.
• B-5: Verified reliability, effectiveness, and security throughout the system's lifecycle.
• B-6: Safety mechanisms to prevent malfunctions or serious failures.
• B-7: Compliance with all relevant laws in system operation.

Encouraging Private Sector Participation

Japan’s Defense Minister, Nakatani Gen, highlighted that these guidelines provide clarity and predictability for companies interested in contributing to defense AI R&D. By setting clear standards and risk management procedures, Japan aims to foster collaboration with private sector innovators while upholding strict ethical and legal standards.

Japan’s Commitment to Responsible AI in Defense

These guidelines align with Japan’s broader AI strategy, detailed in the July 2024 “Basic Policy for Promoting AI Utilization,” which underscores a "Responsible AI" approach. Key focus areas include target acquisition, data analysis, command and control, logistics support, unmanned assets, cybersecurity, and administrative efficiency.

Japan’s Role in International AI Governance

At the United Nations, Japan has advocated for a ban on lethal autonomous weapons systems without human control. The UN Secretary-General has called for a legally binding instrument by 2026 prohibiting such weapons.

While some countries remain reluctant to engage on this issue, Japan continues to push a human-centric framework and collaborates with allies through initiatives like the U.S.-led “Political Declaration on Responsible Military Use of Artificial Intelligence and Autonomy.” This declaration stresses adherence to international law, including IHL, in military AI applications.

Conclusion

Japan’s new AI guidelines for defense R&D, combined with the recent AI legislation, aim to strengthen the country’s AI capabilities while ensuring responsible use in defense systems. Clear risk management and compliance processes reassure businesses and reinforce Japan's commitment to international norms and ethical standards.

For professionals involved in AI development, legal compliance, and management within defense or related sectors, understanding and integrating these guidelines is crucial. They offer a practical framework for balancing innovation with ethical responsibility and legal accountability.

For further learning on AI applications and governance, consider exploring courses on Complete AI Training.