Recall
The proportion of true positive predictions to all actual positive instances in a dataset, measuring a model's ability to identify all relevant instances. Also called sensitivity.
In plain language
Of all the actual positives, how many did the AI find? If there were 100 spam emails and the filter caught 80 of them, its recall is 80%. A high recall means the AI is good at catching the thing you are looking for.
Why this matters
Recall is a critical metric in high-stakes AI applications where missing positive cases carries significant consequences. Your AI governance framework should specify minimum recall thresholds for systems used in risk assessment, fraud detection or compliance monitoring.
Relevance
ImplementationModel performance metrics like recall must be monitored continuously; organisations deploying AI in regulatory or safety-critical contexts need documented recall baselines and alert thresholds to ensure the system catches required cases.
Related terms
Putting recall into practice in your organisation?
Ready to transform your AI strategy?
Partner with Australia's AI strategy and governance specialists. From adoption roadmaps to ISO 42001 audit readiness.