FRAMEWORK · VERSION 1.0

AI-IRF

AI Incident Response Framework — operational response architecture for AI systems.

Operational AI incident response

AI-IRF translates AI-specific incident risk into repeatable preparation, detection, classification, containment, recovery, notification and learning practices.

19Domains
79Controls
1.0Current release

Use the publication

Documentation Center

Public specifications, FAQ, sector notes and access boundaries.

79-control catalogue

Search the canonical public AIIS control records.

AIIS schemas

Schema, profiles, playbooks, severity and control-record JSON.

Publication status

Authoritative count

AI-IRF v1.0 defines 79 controls across 19 domains. The AIIS machine-readable control register is the authoritative structured source. Test vectors and automated conformance objects are planned for a future release.

Framework libraries

HTML FAQ

Published answers on adoption, operations, evidence, status and licensing.

Framework crosswalks

Published AI-IRF relationships with UAIF and GAISSF.

3 sector-guidance notes

Financial services, healthcare and critical infrastructure.

Research library

Response preparation, containment, recovery and learning research.

Related methodology

AI Investigation Readiness — Supporting operational methodology for preparing organisations to use AI during investigative activities. It is not a fourth framework, certification programme or assessment standard.

Launch announcement

Read the GAISSF Ecosystem launch announcement and Founder Note.