Your transaction monitoring alerts are over ninety per cent false positive. Is that a defect, and what would you do about it?

Short answer

It is mostly not a defect. Monitoring is a screening control tuned for recall against a base rate so low that high precision is arithmetically impossible, so the work is measuring the rate honestly, moving volume from rules to models and from analysts to enrichment.

This is an AML transaction monitoring tuning problem. Good AML work separates model threshold, scenario design, data quality, typology coverage, and investigator capacity before lowering alerts. Transaction monitoring is useful only if AML risk remains visible after false positives fall. Weak AML answers celebrate a lower transaction monitoring queue without proving that suspicious behaviour is still being caught.

AML tuning should leave a defensible sample trail for the alerts that disappeared.