GovernanceCore
Operations

AI incident

Working definition

Reviewed 30 July 2026

An AI incident is an event, circumstance, or series of events in which the development, use, or malfunction of one or more AI systems directly or indirectly leads to harm to people, rights, property, communities, critical infrastructure, or the environment.

Context

Why it matters

Consistent incident definitions make escalation, regulatory reporting, root-cause analysis, and cross-system learning possible.

Operating note

What this looks like in practice

  1. 01Define reportable events and near misses before launch, not during a crisis.
  2. 02Capture system version, inputs, outputs, affected groups, controls, and decision logs.
  3. 03Connect incident findings to model changes, control improvements, and risk reassessment.

Sources & further research

Primary authority anchors the definition. Research links add conceptual or operational depth. External sources may update independently; always verify legal duties against the current official text.

  1. Official source

    Overview and methodology of the AI Incidents and Hazards Monitor

    Provides interoperable definitions for AI incidents and AI hazards and explains the OECD monitoring method.

    OECD.AI
    2024    Open source ↗
  2. Official source

    Artificial Intelligence Risk Management Framework 1.0

    Treats post-deployment monitoring and response as continuous risk-management activities.

    NIST
    2023    Open source ↗