Using Compliance co-pilot to evaluate AML hits

Last updated: May 14, 2025

Compliance Copilot is designed to simplify the process of reviewing AML (Anti-Money Laundering) hits by sharing a summary and assigning a relevance score to each result, helping compliance officers quickly identify whether a hit warrants deeper investigation.

Once you run AML Screening on a user, scroll down to the AML section and follow the below steps.
Example in the video: We ran a AML check on Donald Trump (DoB 1999, United Kingdom) which returned 4 hits.

  1. Click Compute Relevance Scores

    • Co‑Pilot analyses each hit and assigns a Relevance Score.

    • A Summary of the hit is also generated with information relevant to the hit

    • You’ll also see quick badges for Name Match, Age Match, and Location Match.

  2. Interpret the Scores

    Relevance Score

    Meaning

    Typical Action

    High

    Strong match across name + at least one other attribute

    Review the document in detail

    Medium

    Partial match (e.g., name only)

    Review if the case is high‑risk; otherwise resolve

    Low

    Minimal or no contextual match

    Mark as false positive and resolve

  3. Resolve or Escalate

    • High Score → open the linked article / watchlist entry, confirm by reading the co-pilot summary, then approve or escalate.

    • Low / Medium → You can confidently Resolve these hits, if the score is low or medium or double check to be certain it’s not your user (e.g., same name, different age/location).


Why Use Compliance Co‑Pilot?

Traditional Review

Co‑Pilot

Manually open every hit (10‑15 min each)

Relevance Scores reduce review times to an average of a minute

Risk of missing context or bias when documents are in a different language or very long

Clean and relevant summaries of the entire hit in English or language of choice

Prone to Human error

Consistent, data‑driven scoring with less than 0.1% error rate.