AI incidents and incident response
Handling events where an AI system causes or nearly causes harm.
AI incidents include harmful or biased outputs, privacy leaks, model failures, attacks and misuse. Response needs clear criteria for what counts as an incident, the ability to switch the system off or fall back, root cause analysis, notification duties and lessons learned.
On the exam
Ask whether the organisation can quickly disable or roll back a model, and whether AI incidents are defined and reported.
Related terms
- Incident responsePreparing for, handling and learning from security incidents.
- Model monitoringWatching a live model’s inputs, outputs and performance.
- Change management for AIControlling changes to models, data, prompts, thresholds and settings.
AAIA Test yourself on this with timed AAIA practice questions in the AAIA Exam Simulation Hub, or read the free AAIA exam guide.
Test yourself with today’s question of the day.
