For Heads-of · Practitioner
Incident Response
A structured process for detecting, assessing, containing and resolving adverse events involving AI systems, such as model failure, a security breach or a fairness harm.
- incident
- response
- crisis
- recovery
Phases
Detection, triage, containment, root cause analysis, remediation, communication and post-incident review.
Key decisions
Does the incident warrant immediate rollback or switching the system off? Who needs to be told, inside and outside the organisation? Which external reporting obligations apply: personal data breach notification, sector regulators and, for high-risk systems, serious incident reporting under the EU AI Act? The AI incident entry sets out which events count and the reporting deadlines.
Playbooks
Responses to the common scenarios should be written and rehearsed in advance: model degradation, a data breach, a discriminatory decision detected, a suspected adversarial attack. Scenarios nobody has rehearsed are the untested AI failure scenarios risk; AI incident response and rollback is the control.
Related Terms
AI Incident
An event in which the development, use or malfunction of an AI system directly or indirectly leads to harm to people's health, critical infrastructure, rights, property, communities or the environment. An AI hazard is an event that could plausibly have led to one. Reporting regimes in the EU and several US states now attach legal deadlines to the most serious cases.
Audit Trail
A complete, timestamped record of the actions an AI system took and the decisions made with it, so that they can be reviewed, explained and attributed afterwards.
Model Monitoring
Continuous observation of a deployed model's inputs, outputs, performance and behaviour, to detect degradation, drift, misuse or security anomalies.
Model Versioning
The practice of recording which model, trained how and on what, was deployed when, enabling comparison, rollback and reproducibility. For systems built on models the organisation did not train, it extends to versioning the surrounding assembly, since that is the only part it controls.
Performance Regression
A decline in model performance over time, typically caused by shifts in the data, concept drift or changes in the environment the model operates in.