More matches found. Less to review.
AI screening that searches wider, assesses in context and reasons like an experienced analyst.
First AML's screening agents run the full screening lifecycle: finding the matches, assessing them, and re-checking them as the case ages.
The search agent goes beyond the data provider to surface results that were missed. The review agent assesses each one against the entity profile already on file, clears what the evidence rules out and flags what it can't settle. When ongoing monitoring produces a new hit, the review agent activates and notifies the user.
Go wide!
Catch more than the data providers
The search agent goes beyond traditional screening data, using web search and paywalled sources to surface possible matches the data provider missed. The advanced search agent brings harder-to-find information into the screening process.
What's in a name?
Contextual decisions, not just a name match
The review agent assesses each hit against the full entity profile, using verified identity, ownership, jurisdiction and documents to distinguish genuine matches from false positives.
Decisions are explained in plain English with sources cited. Anything the evidence can't settle is flagged for human review.
Come again?
Automate ongoing monitoring reviews
When ongoing monitoring produces a new hit, the review agent activates and runs against the latest available information. New results can be surfaced for a quick batch review, or automatically accepted where the firm's risk settings allow it.
"The AI screening agent is a massive help, particularly as it provides reasoning as to why certain results are false matches."
Frequently Asked Questions
How do you train the agent?
The agent is not “trained”, but has been built using our reasoning engine, which includes rules, prompts and intelligence written by our team, encoding how an experienced compliance analyst would approach a hit. No customer data is used for this process.
Is my data secure and where is it processed?
Yes, your data is secure. It runs inside First AML's existing secure environment on AWS, Sydney for ANZ (Melbourne as backup) and Ireland for the UK (Frankfurt as backup), under the same security controls, access rules and audit logging as the rest of the platform.
How should I use it and when should I still rely on my analysts?
It clears the obvious noise so your analysts spend their time on the hits that matter; it doesn't replace their judgement. We recommend you read the reasoning on true matches and, before accepting a batch, spot-check a sample of false matches rather than accepting all and treat "review required" hits as you do today. The compliance decision stays with you.
How accurate is it?
We've tested it against 5,500 hits from 400 cases globally, without showing it the answer, then compared its call with the human one. It resolves around 9 in 10 hits on its own, and where it clears a hit as a false match, humans agreed 99.9% of the time. Anything it isn't sure about goes to a human for analysis.
Can I see why it made the decision?
Yes. Every decision comes with a plain English explanation of the reasoning, with the underlying evidence and sources attached. The decision is clearly marked as AI, so there's a record of what it decided, why it decided it and what information it relied on.
What happens when it isn't sure?
It doesn't guess. Hits the agent can't confidently clear are flagged for human review, with the reasoning and evidence shown alongside them. Your analysts can see exactly what the agent found and why it couldn't settle the match.
Why firms trust First AML
Built on real AML experience
We processed over 2 million AML cases ourselves before building the platform - so it’s shaped by what works in practice.
Local support that understands compliance
Our Customer Success team is based in-market and works alongside local firms and consultants to help turn policies into day-to-day workflows.
Solutions, not sales pitches
We take a solution-led approach, configuring First AML around your risk framework, operating model and regulatory context.
See First AML in action.
The only AML platform you need to run world-class onboarding.