Japan AI Basic Plan (second plan, July 2026)
Issued by
Cabinet of Japan
- September 30, 2026 · Correction — Corrected the adoption date, legal basis, and four directions, and noted its replacement by the 14 July 2026 plan (fact-check finding). (Cody Maxwell)
- October 1, 2026 · Correction — Updated the entry to describe the second AI Basic Plan, adopted on 14 July 2026, which replaced the December 2025 plan. (Cody Maxwell)
- October 1, 2026 · Correction — Rewrote the practical steps, newsletter hook, search description, and audience fields to match the entry's corrected content. (Cody Maxwell)
Japan's Cabinet adopted its second AI Basic Plan on 14 July 2026, replacing the first plan of December 2025. It responds to the rapid rise of agentic AI and puts domain-specific 'vertical AI' and physical AI at the centre of Japan's strategy. It directs government policy and places no direct duties on companies.
Applies To
Overview
Japan's Cabinet adopted the second Artificial Intelligence Basic Plan on 14 July 2026, subtitled 'Japan AX: stronger and richer'. The government must prepare such a plan under Article 18 of the Act on Promotion of Research and Development and Utilization of AI-Related Technologies (Act No. 53 of 2025). It replaces the first plan, adopted on 23 December 2025, which the Cabinet Office now lists as a past plan. The plan says AI has moved quickly from tools that help with documents to agentic AI that plans, acts, and checks its own work. Japan's strategy now centres on vertical AI built for specific industries and physical AI that acts in the real world. It also stresses open 'AI sovereignty', meaning less dependence on any single country or company. It notes that about 180,000 government staff already use the government AI service 'Gennai'. It keeps four principles: balance innovation with risk, learn by trying, respond with agility, and align domestic and international policy. It also keeps the first plan's four directions: use AI, create AI, raise AI trustworthiness, and work with AI. On governance, it calls for agile oversight that combines rules, technical standards, and organisational practice. It strengthens the AI Safety Institute, including guidelines for agentic AI and sharing of AI vulnerability and incident information. It also commits to rules for government use of AI agents and continued review of the AI Act. Japan will keep leading international work such as the Hiroshima AI Process. The plan directs government policy and encourages voluntary action by companies, but sets no duties of its own.
Key Requirements
- •A statutory government plan under Article 18 of Japan's 2025 AI Promotion Act; it sets no duties for private companies.
- •Replaces the first AI Basic Plan of 23 December 2025.
- •Focuses on agentic AI, vertical AI for specific industries, physical AI, and open AI sovereignty.
- •Keeps four directions: use AI, create AI, raise AI trustworthiness, and work with AI.
- •Strengthens the AI Safety Institute, including guidelines for agentic AI and incident information sharing.
- •Commits to rules for government use of AI agents and ongoing review of the AI Act.
What Your Organization Must Do
- →Read the second AI Basic Plan to see where Japanese government AI funding and policy attention will go.
- →Tell leadership this plan directs government policy and creates no new duties for your company.
- →Track AI Safety Institute output, especially guidelines for agentic AI that plans and acts on its own.
- →Watch the promised review of Japan's 2025 AI Promotion Act for any future binding obligations.
- →Map products against the plan's focus on vertical AI for specific industries and physical AI.
- →Note planned rules for government use of AI agents if you sell to Japanese public bodies.