Overview
When an Ongoing Monitoring alert is ready for review, Legl's AI-assisted review brings together the evidence from your checks and carries out its own research across the open web. It then presents a single recommendation on whether the flagged record relates to your client, along with the reasoning and the sources behind it.
The aim is to give your compliance team a clearer, faster review. The recommendation is a decision-support aid only: a reviewer still confirms the outcome and marks the alert as reviewed.
How this works
Each time an alert is opened for review, the agent works through the available evidence and produces a structured summary. It uses the details you already hold on the client, checks the source supplied with the alert, and searches the open web for corroborating information, then sets out a recommendation with a confidence level.
The recommendation and confidence
At the top of the review, the agent answers a single question: is this your contact? It gives a recommendation, such as a likely match or a likely different person, together with a confidence level and a short reason summarising what drove the assessment.
What was flagged and what matched
The review separates two things. What was flagged is the reason the alert was raised, for example a politically exposed person classification. What actually matched is the specific detail that aligns between your client and the flagged record, such as a name or an address.
Full comparison
A side-by-side comparison shows what you hold on your client against the flagged record, field by field. Each row shows a result, such as a match or no data available, and where a detail was found it links out to the source.
How the agent investigated
The review shows the research behind the recommendation in two parts:
Source review: the agent checks the source supplied with the alert and assesses whether it connects to your client.
Web search: the agent searches the open web beyond the supplied material to test whether your client and the flagged entity are the same person. The searches it ran are listed and can be opened, and it records where nothing relevant was found.
ℹ️ Further guidance
What triggers this
AI-assisted review runs on Ongoing Monitoring alerts for individual clients, so the evidence and recommendation are ready when you open an alert to review it. No action is required to trigger it.
What you will see
When you open an alert, the review page shows:
A summary of the client and the alert, including the current status, when it was created, who reviewed the initial report, the date of the last CDD check, and whether a PEP or sanction has been confirmed.
The recommendation banner answering whether this is your contact, with its confidence level.
The evidence sections: what was flagged, what matched, the full comparison, and how the agent investigated.
The actions available on the alert, including Assign reviewer, Download PDF, and Stop monitoring.
To record your decision, add a review comment, mark whether the flagged record is or is not your client using Match or Not a match, and then select Mark as reviewed. For the full step-by-step, see the guide on reviewing a monitoring alert.
Important information
AI-assisted review is a decision-support aid. Your firm must still apply its own risk-based judgement to every alert.
Alerts are not closed automatically. A reviewer must confirm Match or Not a match and select Mark as reviewed.
A match against a PEPs, sanctions, or watchlist source does not confirm that your client is on that list. Always review the full details before taking action.
The sources checked and the searches run are recorded, giving you a full audit trail you can open and verify.
The recommendation summarises the available evidence. It is informational only and does not constitute legal or compliance advice.