Every AI Policy Ever · open in the interactive register

Published (non-binding, current)Non-binding · Guidelines / framework

Approach Book for AI Incident Response (AI-IRS): A Foundation for Trustworthy AI

AIインシデントレスポンス・アプローチブック

Record JP-07

Issued by
Japan AI Safety Institute (J-AISI)
Jurisdiction
Japan
Level · region
National · East Asia
Published / adopted
2026-01-09
Language
Japanese, English

Summary

Framework of the Japan AI Safety Institute for responding to incidents caused by risks unique to AI systems, for executives such as Chief AI Officers and CISOs, security teams (CSIRTs, SOCs) and policymakers. Starting from the premise that AI incidents are inevitable, it stresses detective controls that minimise damage and presents the AI Incident Response System (AI-IRS), which adds observability (detection and analysis) and controllability (containment, eradication and recovery) to existing incident response. It analyses RAG systems and AI agents, links the approach to SBOM for AI and the NIST Cyber AI Profile, and calls for transparency and traceability across the AI supply chain.

Safety topics

incident reportingsafety instituteaccountability & liability

Status notes

Published in Japanese and English, each with a summary edition. The book presents a conceptual framework that does not replace existing information security incident response frameworks such as NIST SP 800-61; implementation guides and procedures are to follow as related deliverables. It also discusses contractual 'action' by AI with reference to the UNCITRAL Model Law on Automated Contracting, sustained budgets for AI incident response, and gives a confirmation flow for the CAIO in an appendix.

Sources

Open the official source https://aisi.go.jp/activity/activity_security/260109/

Local copy

Archived. Last attempt 2026-10-18 00:45:37 UTC

Official source (third-party copy, not published) archive/_runs/2026-10-14-discover-leads-2026-10-14/C012/2026-10-14/official.html
Retrieved 2026-10-14 13:10:49 UTC
SHA-256 11528fe0ca3922ee65a376662fc4890585527502a3e3e2877c03642b19144ebd
Linked document (PDF) (third-party copy, not published) archive/_runs/2026-10-14-discover-leads-2026-10-14/C012/2026-10-14/document-1.pdf
Retrieved 2026-10-14 13:10:53 UTC
SHA-256 dc5ed3714b7f05f1eed8e3c56f7a807d085b0c230978eb7ab2e41acf28202b50
Linked document (PDF) (third-party copy, not published) archive/_runs/2026-10-14-discover-leads-2026-10-14/C012/2026-10-14/document-2.pdf
Retrieved 2026-10-14 13:11:00 UTC
SHA-256 5d613a795625e0a52fcfa1ea2c3a08a63b2c9e3e2ab542bf2a159c08d7b624d8

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 bilingual release page on aisi.go.jp, J-AISI's English edition of the Approach Book (42 pages, dated January 9, 2026) and its English summary edition (January 2026), all copies of 14 Oct 2026 that the archiver re-checked unchanged on 18 Oct 2026; and the Japanese edition (ai-irs_v1.0_ja.pdf, 38 pages, stored 18 Oct 2026), titled AI インシデントレスポンス・アプローチブック and dated 令和８年１月9日 (9 Jan 2026), issued by AI セーフティ・インスティテュート. They confirm the title, the original title, the release on 9 January 2026, the issuer and the content; the summary and notes were checked against the English edition and still hold. The two editions have the same structure and date, and the English edition carries no translation notice, so both are treated as J-AISI's own texts (english_version 'authentic'); neither says which one prevails.
Second check: Independent check by Claude Code (claude-sonnet-5-5), 2026-10-18: 6 field(s) supported.

History

  1. 2026-10-18Re-verified against its sources, no change — pipeline run 2026-10-18-recheck-east-asia-1 (recheck), candidate JP-072026-10-18-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 C012 (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). Approach Book for AI Incident Response (AI-IRS): A Foundation for Trustworthy AI. 2026-01-09. https://aisi.go.jp/activity/activity_security/260109/. Status: Published (non-binding, current); verified 2026-10-18 (level A). Local copy archive/_runs/2026-10-14-discover-leads-2026-10-14/C012/2026-10-14/official.html retrieved 2026-10-14 13:10:49 UTC, SHA-256 11528fe0ca3922ee65a376662fc4890585527502a3e3e2877c03642b19144ebd. Every AI Policy Ever, version 1.18 (snapshot 2026-10-18), record JP-07.

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.