Plan
AI Incident Response Plan
This plan defines the AI-specific incident response path that operates alongside the security and privacy incident procedures. It covers prompt injection, indirect prompt injection, jailbreak, data poisoning, model inversion, membership inference, retrieval leakage, agent compromise, unsafe tool execution, hallucination causing material harm, model degradation, and supplier model incidents.
- Type
- Plan
- Version
- 1.0.13
- Date
- 2026-10-13
- Owner
- AI Security Maintainer
- Approving authority
- Governance Library Maintainer
- Classification
- Public
- Confidentiality
- Public
- Category
- AI Governance
- Review frequency
- Annual and upon material AI threat-pattern, regulation, or system-architecture change
- Licence
- CC BY-SA 4.0
Read the full text on GitHub (opens in a new tab)
Outline
- Purpose
- Scope
- AI incident classes and triggers
- AI incident reporting and disclosure channels
- Severity criteria
- Lifecycle
- Coordination with other streams
- Evidence requirements
- Framework alignment
- Limitations
Frameworks it maps to
- OWASP LLM Top 10 (2026)
- MITRE ATLAS
- NIST AI RMF Generative AI Profile (NIST AI 600-1)
- NIST AI 100-2e2025
- ISO/IEC TR 27563:2023
- ISO/IEC 42001:2023
- EU AI Act
- ISO/IEC 27035-3:2020
- SANS CAISG v1.4
- OWASP AI Exchange
- ETSI EN 304 223 V2.1.1
Related documents
- AI Governance and Risk Frameworkai
- AI Security and Risk Standardai
- AI and Agentic Development Security Standardai
- AI Adversarial Test Referenceai
- AI System Register Templateai
- Security Incident Response Proceduresecurity
- Data Protection and Privacy Breach Response Procedureprivacy
- Cross-Domain Incident Coordination Procedureresilience
Next
All Documents