False positive (screening)

A false positive in screening is an alert that flags a customer, owner or counterparty as a potential match to a sanctions list, PEP database or adverse media result when, on review, the person or company is not the one listed. The opposite error, a false negative, is a real match the screen fails to raise.

Why it matters in KYB and fraud review

Screening tools match names, and names are not unique. Common names, transliterated names, short company names and weak aliases all generate hits that an analyst has to clear by hand. How those hits are cleared decides two things at once: how fast legitimate customers get through, and whether the rare true match is noticed.

The failure modes run both ways. Clearing too fast, by habit, lets a true match through. Escalating everything buries the compliance team and delays customers for no gain.

What an analyst checks

OFAC's guidance on assessing name matches sets out the method that applies to most screening: "compare all of the details in the OFAC listing with the information available to you," including full name, aliases, nationality, passport numbers, date of birth, addresses and business registration numbers. It suggests asking whether only the first or last name matches, and whether other identifiers differ.

A good clearance:

• Compares at least one strong identifier beyond the name, such as date of birth, nationality, registration number or address.
• Checks the listing's aliases, including weak aliases, before relying on a name difference.
• Considers context: a 30-year-old founder of a local bakery is unlikely to be a listed 70-year-old official, but record the fields compared rather than the intuition.
• Records the decision and its basis so a second reviewer or auditor can follow it.
• Escalates when identifiers are missing or conflicting, rather than guessing.

If there is an exact match or a close match with multiple similarities, OFAC's guidance says to follow your sanctions compliance procedures (see alert escalation).

Common pitfalls

• "Different middle initial, cleared." Names alone rarely settle a match.
• Clearing a hit on the entity but not rechecking when a new owner or director is added.
• Suppressing recurring hits permanently, so a later genuine listing of the same person is missed.
• Clearance notes that say "not a match" without saying why.

Related terms

OFAC SDN list, adverse media screening, politically exposed person, alert escalation.

Sweat AI is an AI-native BPO for banks and fintechs, working onboarding and fraud queues 24/7. Our analysts clear screening hits by comparing every available identifier, write down the basis for each decision, and escalate true and unresolved matches to your team. See fraud alert review.