The Guide the Vendors Don't Write

The guide the vendors don't write

Search "reduce AML false positives" and you'll find dozens of guides. Every one of them is written for whoever buys screening software, tighter thresholds, better data, smarter algorithms, fewer alerts overall. None of them are written for the analyst who already has an alert open right now and has to decide, correctly, defensibly, in the next few minutes, whether it's real.

That's a different skill, and it's the one this guide actually teaches. Most of an analyst's real workload isn't catching the rare genuine hit, it's disposing everything else properly, and getting that wrong in either direction, clearing a real match or escalating a harmless namesake, has real consequences.

How to use this guide

Read the four patterns first. Then work each scenario, make your call, and read why, both for the right answer and every wrong one. The knowledge check at the end pulls the same patterns from new angles. Do not skip ahead.

Four Patterns

Four patterns, four visual memory cards

Most false positive misjudgments trace back to one of these four shapes. Learn to recognise them before the scenarios test them.

👥
Visual Memory: The Crowd
A name alone, in a country of millions, means almost nothing. Picture a crowded room where everyone shares a surname. A match on name alone is just confirming you're in the right room, it tells you nothing about which person in it you're looking at. The more common the name, the less a bare match is worth, and the more the secondary identifiers, date of birth, nationality, address, associated entities, have to do the actual work of confirming or ruling out identity.
🎂
Visual Memory: The Wrong Birthday
A DOB mismatch is powerful, but not automatically conclusive. A clean date of birth mismatch feels like a definitive clear, and often it is. But sanctions and PEP records are frequently incomplete, a DOB field left blank, a year-only record, or a format inconsistency between systems (day-month-year versus month-day-year) can produce a false mismatch that looks conclusive but isn't. The birthday has to be genuinely present and genuinely reliable on both sides before it clears anything.
🔤
Visual Memory: Same Name, Different Alphabet
Transliteration produces multiple correct spellings of the same real name. A name originating in Arabic, Cyrillic, or Chinese script doesn't have one correct English spelling, it has several equally valid ones depending on the transliteration convention used. A spelling variant between your customer record and a sanctions entry isn't evidence of a different person, it can be the same person rendered through a different transliteration standard. Treating spelling variance as automatic grounds to clear is exactly as wrong as treating it as automatic grounds to escalate.
📰
Visual Memory: The Wrong Person in the Headline
Adverse media flags the name, not necessarily the person. A negative news hit returns because a name appears in an article, not because the system has confirmed the article is about your specific customer. A common name attached to a serious headline is one of the easiest ways to escalate the wrong person entirely, or worse, clear the right one because the obviously wrong headline made the whole alert look like noise.
🪪
Scenario 01 · Sanctions name match
The Same Name
The Crowd The Wrong Birthday
Alert Evidence
🔔
Alert triggered
Mohammed Al-Rashid matches a sanctioned individual by name
→
check DOB
🎂
DOB checked
On file, doesn't match the sanctioned record
→
verify reliability
⚠️
Under review
No other identifiers populated to cross-check
⚖️
What do you do? Make the call

Your screening system flags a customer named Mohammed Al-Rashid against a sanctions entry for an individual with the same name, designated for facilitating sanctions evasion. Your customer's date of birth is on file and doesn't match the sanctioned individual's recorded DOB. No other identifiers, nationality, address, or associated entities, are populated on the sanctions record to compare against.

🔤
Scenario 02 · Transliteration
The Spelling Variant
Same Name, Different Alphabet
Alert Evidence
🔔
Alert triggered
Yusuf Al-Bakri vs. sanctioned "Yousef Al-Bakry"
→
check identifiers
🌍
Nationality matches
No DOB populated on the sanctions record
→
assess spelling
⚠️
Under review
Spelling consistent with transliteration, not a different name
⚖️
What do you do? Make the call

A customer named Yusuf Al-Bakri is flagged against a sanctions entry for "Yousef Al-Bakry," a similar but not identical spelling, same underlying name transliterated differently from Arabic script. Nationality on both records matches. No DOB is populated on the sanctions record at all.

Screening a customer or entity? Run a free sanctions, PEP, and adverse media check with adjustable fuzzy matching, no account required.
Screen an entity →
📰
Scenario 03 · Adverse media
The Headline
Wrong Person in the Headline
Alert Evidence
🔔
Alert triggered
James Whitfield named in a fraud sentencing headline
→
check details
📍
Details compared
Article subject: 34, different region of the country
→
compare subject
⚠️
Under review
Your director: 58, same address over a decade
⚖️
What do you do? Make the call

An adverse media screen on a new corporate customer's director, James Whitfield, returns a news article headlined "James Whitfield Sentenced for Fraud." The article describes a James Whitfield convicted in a regional court for a fraud scheme, aged 34, based in a different part of the country from your customer. Your director is 58 and has lived at the same address for over a decade.

Knowledge Check
Five questions. How well do you dispose an alert?
1. What does "The Crowd" pattern warn against?
2. Why can a DOB mismatch sometimes be unreliable evidence for clearing an alert?
3. A name spelled differently between your customer record and a sanctions entry, both from the same script origin, most likely indicates:
4. What's the actual problem with disposing an alert as "common name, not relevant" with no further detail?
5. What's the central skill this guide actually teaches?
0/5
Quick Reference

At a glance

Four patterns, the trap that makes each one look routine, the tell that actually gives it away, and the response that fits.

PatternThe trapThe tellResponse
The Crowd Name match feels significant alone Common name, no secondary identifiers Verify DOB, nationality, or other identifiers before deciding either way
The Wrong Birthday A mismatch looks automatically conclusive DOB field blank, partial, or format-inconsistent on one side Confirm both DOBs are genuinely complete and comparable first
Same Name, Different Alphabet Spelling variance looks like proof of a different person Names from the same script origin, transliterated differently Treat variance as consistent with the same name, assess other identifiers
Wrong Person in the Headline Serious headline triggers reflexive escalation Age, location, or role details don't match your subject State the specific distinguishing facts, not a general "probably unrelated"
State the Evidence, Not the Feeling

Every disposition rests on the same discipline.

State the specific evidence, not a feeling. A file that says "cleared, DOB confirmed populated and mismatched on both records" survives review. A file that says "looks fine" doesn't. FinCrimeRadar's screening tool surfaces exactly the identifiers this guide teaches you to weigh, sanctions and PEP data with adjustable fuzzy matching, free to use.

Ready to try it on a real name? Free. No account needed. Sanctions, PEP, and adverse media, with adjustable fuzzy matching.
Screen an entity →