World Watch/Japan/Artificial Intelligence

Artificial Intelligence · Japan

Artificial Intelligence - Japan

Comprehensive lawAct on Promotion of Research and Development and Utilization of AI-Related Technologies ('AI Promotion Act', 2025), administered by the AI Strategy Headquarters within the Cabinet, complemented by the non-binding METI/MIC AI Guidelines for Business.

Japan enacted its first national, generally-applicable AI law in May 2025, but it is a deliberately light-touch 'innovation-first' framework law that sets principles and coordinating machinery rather than detailed obligations, prohibitions, or penalties. Governance combines this statute with soft-law guidelines and a Cabinet-level AI Strategy Headquarters; Japan continues to favor voluntary compliance over EU-style hard regulation. A first national Basic AI Plan was adopted by the Cabinet in December 2025.

First national AI law

The AI Promotion Act passed the Diet on 28 May 2025; most provisions took effect 4 June 2025 and it came into full effect later in 2025, making Japan one of the first major economies with a dedicated, economy-wide AI statute.

No fines or bans

The Act imposes no monetary penalties, mandates, or use bans; authorities may only give advice, request information, or publicly name non-compliant actors. It encourages voluntary compliance rather than creating enforceable duties.

AI Strategy Headquarters

The Act establishes a Cabinet-level AI Strategy Headquarters chaired by the Prime Minister (with all ministers as members) to coordinate a whole-of-government approach and formulate the national Basic AI Plan.

First Basic AI Plan (Dec 2025)

The Cabinet adopted Japan's first national Basic AI Plan on 23 December 2025, framing 'reliable AI' that balances innovation with risk management and backed by a planned multi-year, ~¥1 trillion support scheme for domestic AI from FY2026.

AI Guidelines for Business

Non-binding AI Guidelines for Business (Version 1.1, issued jointly by METI and MIC on 28 March 2025) set foundational values and ten cross-sector principles (fairness, privacy, safety, transparency) using a risk-based, agile-governance approach.

Focus on 'high-impact' frontier models

Implementing guidance issued 19 December 2025 emphasizes risk-based lifecycle governance, with attention to developers of 'high-impact' frontier AI models; a precise technical definition is expected from an Expert Investigation Team in Q3 2026.

Machine-assisted translation · verified 5/23/2026 · orientation, not legal advice. English version →