Skip to content

CS Professional · IFSCA - Regulations, Listing and Compliances · Overview of FinTech and Service Providers in IFSC

Nirmaan Bank's IFSC banking unit uses a machine-learning model to flag unusual transaction patterns for anti-money-laundering review. A compliance officer notes that the model sometimes flags genuine customers. What is the most appropriate step consistent with sound governance of AI in compliance?

The bank should keep human oversight of flagged cases and periodically validate and tune the model to reduce false positives. Automated tools are not error-free, and monitoring duties and record-keeping must continue. Removing review, deleting records or suspending monitoring would weaken anti-money-laundering compliance.

  1. ASwitch off human review, since the model is automated and therefore accurate
  2. BRetain human oversight of flagged cases and periodically validate and tune the model for false positivesCorrect
  3. CDelete all flagged-case records to avoid evidence of errors
  4. DStop monitoring altogether until the model has no errors

Explanation

AI monitoring tools generate false positives, so entities should keep human review, validate the model and calibrate it regularly. Removing human review raises risk, deleting records breaches record-keeping duties, and stopping monitoring would leave AML obligations unmet.

Did you get it right without looking?

One question tells you little. A timed set on Overview of FinTech and Service Providers in IFSC shows your real accuracy, how long you take and where you lose marks.

Practise Overview of FinTech and Service Providers in IFSCFree diagnostic mock

More Overview of FinTech and Service Providers in IFSC questions