Napier AI has weighed in on one of the thorniest debates in financial crime compliance: the gap between what artificial intelligence can theoretically do for transaction monitoring (TM) and what institutions are actually achieving today.
The discussion, raised during a recent ACAMS New Jersey chapter session, centred on a widening divide between AI used for efficiency and AI used for genuine detection.
According to Napier AI, most institutions currently deploy AI to help investigators clear alert queues faster or draft Suspicious Activity Report (SAR) narratives, rather than to catch anything the underlying rules missed. A smaller cohort is now pushing AI onto the detection side itself, running models alongside rules to learn customer behaviour and surface activity that static thresholds were never built to catch, a shift Napier AI links directly to the Financial Crimes Enforcement Network’s (FinCEN) April Notice of Proposed Rulemaking (NPRM).
Napier AI notes that FinCEN has explicitly named AI as a factor in future enforcement decisions, reframing the industry question from “how much time does AI save” to “how much crime does it help identify”. This, Napier AI argues, should push compliance teams to reconsider defensive SAR filing habits, where narratives are overloaded with every transaction rather than telling a precise, investigation-led story.
Napier AI also cautions against mistaking speed for progress. If false positive rates remain above 90%, faster processing of the same alerts does not equate to reduced risk exposure. Alert volumes typically rise in year one of AI-augmented monitoring, not fall, as models surface previously unhandled categories of activity.
On automation limits, Napier AI maintains that AI should not auto-close alerts or file SARs end-to-end for the foreseeable future, since accountability for financial crime decisions cannot be outsourced to a model. Instead, the better fix is tuning detection so fewer low-value alerts are generated in the first place.
Ultimately, Napier AI frames rules and AI as complementary rather than competing: models surface what rules cannot, and once a pattern is validated through real SARs, it gets folded back into an explainable rule, creating a feedback loop regulators are more likely to trust.
For more insights, read the full story here.
Read the daily FinTech news
Copyright © 2026 FinTech Global
