Every AI Policy Ever · open in the interactive register

Published (non-binding, current)Non-binding · Report (government-commissioned)

AMAIS: Activity Map on AI Safety

Record JP-10

Issued by
Japan AI Safety Institute (J-AISI)
Jurisdiction
Japan
Level · region
National · East Asia
Published / adopted
2025-02-07
Language
English

Summary

Discussion paper of the Japan AI Safety Institute presenting the Activity Map on AI Safety (AMAIS), an overview of the activities needed to achieve AI safety, drawn from the outcomes of the Hiroshima AI Process and the AI Seoul Summit. It maps fifteen activities, among them AI safety evaluation and red-teaming, addressing societal risks, responsible information sharing, content authentication and provenance, security throughout AI lifecycles and data quality, against the goals of safety, inclusivity and innovation, and places the work tracks of the International Network of AI Safety Institutes on the map. It also sets out terminology challenges for further discussion.

Safety topics

safety instituteinternational cooperationevaluation & testing

Scope

Borderline scope Existing scope_note kept.

Status notes

J-AISI describes the paper as a living document to promote international discussion on AI safety and names a technical discussion on the map and on terminology as a common language as the next step. For each activity the paper lists its sources in the Hiroshima Process International Guiding Principles and Code of Conduct and in the Seoul Declaration and Statement of Intent, and examples of related terms. J-AISI's Japanese page calls it AIセーフティに関する活動マップ（AMAIS）.

Sources

Open the official source https://aisi.go.jp/output/output_information/250207_1/

Local copy

Archived. Last attempt 2026-10-18 00:45:54 UTC

Official source (third-party copy, not published) archive/_runs/2026-10-14-discover-leads-2026-10-14/C015/2026-10-14/official.html
Retrieved 2026-10-14 13:11:10 UTC
SHA-256 0f31dce36981cc4047ab0a7506e9004f8222e23d420f20e0db6c307189f63892
Linked document (PDF) (third-party copy, not published) archive/_runs/2026-10-14-discover-leads-2026-10-14/C015/2026-10-14/document-1.pdf
Retrieved 2026-10-14 13:11:26 UTC
SHA-256 9f19797df4cadcea4c4ed01239790069114821cd46525cf50388fee3986c81cc

Verification

A
The instrument's own official text, or an official register, was stored and confirms the title, date and status. checked 2026-10-18 · official-url-fetched
Recheck of 18 Oct 2026. Stored and read: J-AISI's release page on aisi.go.jp (Japanese and English; copy of 14 Oct 2026, unchanged on 18 Oct 2026), the paper itself (10 pages, English only, cover 'Discussion Paper', 'Japan AI Safety Institute', 'Feb. 2025') and J-AISI's list of Information outputs (copy of 18 Oct 2026), which dates the release 2025年2月7日. They confirm the title, the issuer, the date, the nature as a discussion paper, the fifteen activities, the sources and terms listed for them, the terminology challenges and the content of the summary and notes. The list of outputs shows no later edition of the map, and nothing stored shows a withdrawal.
Second check: Independent check by Claude Code (claude-sonnet-5-5), 2026-10-18: 5 field(s) supported, fixes applied to scope_note.

History

  1. 2026-10-18scope_note: A discussion paper rather than guidance, which says it does not indicate activit → Existing scope_note kept. — pipeline run 2026-10-18-recheck-east-asia-1 (recheck), candidate JP-102026-10-18-recheck-east-asia-1 · Claude Code (claude-opus-5-5)
  2. 2026-10-14Added to the register — found in pipeline run 2026-10-14-discover-leads-2026-10-14, candidate C015 (JP/ai-bodies)2026-10-14-discover-leads-2026-10-14 · Claude Code (claude-opus-5-5)

Related records

Japan AISI Japan · Published (non-binding, current)

Citation

Japan AI Safety Institute (J-AISI). AMAIS: Activity Map on AI Safety. 2025-02-07. https://aisi.go.jp/output/output_information/250207_1/. Status: Published (non-binding, current); verified 2026-10-18 (level A). Local copy archive/_runs/2026-10-14-discover-leads-2026-10-14/C015/2026-10-14/official.html retrieved 2026-10-14 13:11:10 UTC, SHA-256 0f31dce36981cc4047ab0a7506e9004f8222e23d420f20e0db6c307189f63892. Every AI Policy Ever, version 1.18 (snapshot 2026-10-18), record JP-10.

Improve this record

Know of an amendment, a status change or an error? Report news or an error on GitHub; the form opens with this record filled in.