TABLE OF CONTENTS
- Overview
- Before you start
- What the agent does with each hit
- What you'll see on the screening report
- Hit statuses and what they mean
- Accepting the agent's decisions
- Changing a decision
- Seeing the full history in the activity log
- Configuring the agent
- How the agent decides, and why you can trust it
- Common questions
- Related articles
Overview
The AI screening agent makes working through screening results faster. It takes the first pass on each hit: it reviews the hit, decides whether it's a genuine match, and writes a short reason in plain language. Its decisions show as suggestions until a person accepts or changes them.
- Case screening. Your First AML Specialist reviews every hit manually, following our Standard Operating Procedure, and accepts or changes each suggestion. Your screening results look the same as before.
- Ongoing monitoring. When ongoing monitoring produces new hits, users who review ongoing monitoring hits see the agent's suggestions and reasoning, and accept or change them.
Before you start
- Roles: users who can already review ongoing monitoring hits can accept or change the agent's suggestions. Access does not change.
- Settings: only the First AML team can change the agent's settings. See Configuring the agent.
What the agent does with each hit
For every hit, the agent reaches one of three outcomes:
- False match suggested. The agent is confident the hit is not your customer. For example, the date of birth or country does not match, or a company only shares a generic word like "International Holding Co".
- True match suggested. The agent has found a hit that looks like a genuine match. It shows the evidence.
- Review required. The agent cannot decide with confidence, so it leaves the hit unresolved for a person to review.
The agent only acts when it is sure, so anything genuinely uncertain still reaches a human.
What you'll see on the screening report
- A summary banner at the top of the report showing what the agent did, for example, "AI screening agent resolved 9 of 10 hits".
- A breakdown below the banner showing how many hits it marked as false matches, how many as true matches, and how many require review.
- A status on every hit, showing the decision, whether it has been accepted and who accepted or declined the agent's status recommendation.
Open any hit to see the full Hit details. The agent's decision appears at the top, and its explanation sits in a Reasoning box in plain language.
Hit statuses and what they mean
The icon shows who made the decision: a sparkle for the agent, a person for someone at First AML or on your team. The word suggested means the agent has decided but nobody has accepted it yet.
| Status | What it means |
|---|---|
| True match suggested or False match suggested, with a sparkle icon and an outline | The agent has made a decision. Nothing is recorded against the hit yet. |
| True match or False match, with a person icon | Your Specialist or someone on your team accepted the agent's decision, or changed it. |
| Review required | This hit needs a person to decide. Either the agent could not decide with confidence, or it has been sent back for review. |
Accepting the agent's decisions
This applies to ongoing monitoring hits. On case screening, your Specialist does this.
- Review the summary banner and the breakdown of what the agent decided.
- Open any hits you want to check and read the Reasoning box.
- Select Accept matches.
The word suggested drops off each status, and the icon changes to a person. Hits marked Review required stay in your queue unchanged.
Changing a decision
- Open the hit.
- Use the Set match status to dropdown to change the status.
- Optionally, add a note under Expand match status notes.
Your decision replaces the agent's and is recorded against your name. The agent's Reasoning is not overwritten.
Seeing the full history in the activity log
Every decision on a screening is captured in the activity log:
- On the case. The Activity log tab on a case shows each screening decision.
- On each hit. Open a hit and expand its Activity log to see the history for that hit on its own.
The log shows each change of status, for example, from Review required to False match suggested to False match. Each entry shows who or what took the action (the agent, your Specialist or your team) and the time. Nothing is overwritten.
Configuring the agent
Only the First AML team can change the agent's settings for your organisation. To turn the agent off, contact your Customer Success Manager or email support@firstaml.com. Any decisions the agent has already made stay in the record.
How the agent decides, and why you can trust it
The agent doesn't judge a hit on the name alone. It uses the context First AML already holds on the case: the verified date of birth, the country, the entity structure and ownership, and the documents collected. That context is what lets it tell a genuine match apart from a coincidence.
Every decision the agent makes is written in plain language, attributed to the agent, and kept in the record alongside human decisions. Nothing the agent does is final until it is accepted.
Common questions
What does "suggested" mean on a status?
The agent has made a decision, but it is not recorded yet. Once it is accepted, the word suggested drops off and the icon changes to a person.
Can the agent mark a hit without anyone accepting it?
No. Every suggestion is accepted or changed by your Specialist (case screening) or your team (ongoing monitoring).
The report shows the agent is still working.
The agent hasn't finished this screening yet. The report updates on its own when it's done.
A decision looks wrong.
Open the hit and use Set match status to to change it. The agent's original decision and reasoning stay in the record.
How do I stop using the agent?
Contact your Customer Success Manager or email support@firstaml.com.