The Problem
Why false positives happen
Common names
Many sanctioned individuals have highly common names (e.g., 'John Smith' or 'Mohamed Ali'). Without context, every matching name flags.
Transliteration variations
Names translated from Cyrillic, Arabic, or Han scripts to Latin characters have multiple valid spellings.
Missing context
Screening a name without a Date of Birth or Country prevents the system from ruling out obvious mismatches.
Weak matching thresholds
Legacy systems relying on simple Levenshtein distance or soundex algorithms produce too much noise on short names.
Stale list data
Using outdated sanctions lists means you might flag entities that have been delisted or cleared by regulators.
Partial name overlap
A person whose middle name matches a sanctioned entity's surname might trigger a flag in naive systems.
Best Practices
How to reduce false positives
Provide Date of Birth
DOB is the strongest disqualifier. A match with the same name but different DOB can often be automatically cleared.
Provide Nationality or Country
Including the entity's country of residence or registration allows the matching engine to apply contextual penalties.
Use Entity Identifiers
Whenever possible, screen using national IDs, passport numbers, or LEIs, which provide definitive matches.
Specify Entity Type
Always declare if you are screening a 'person' or an 'entity' to prevent companies from matching with individuals.
Leverage Source Evidence
Use a system that provides per-field confidence breakdowns so your team can quickly understand why a match occurred.
Implement Human Review
Set risk thresholds to automatically clear low-confidence hits, escalating only credible matches for manual review.
Verifex Platform
How Verifex helps manage noise
We design our screening infrastructure to help you confidently reduce false positives while maintaining strict regulatory compliance.
10-penalty pipeline
Verifex applies sequential penalties for mismatches in name structure, DOB, country, and identifiers, significantly reducing noise.
Context-aware scoring
Advanced context-aware matching and scoring signals weigh the statistical probability of a match based on field rarity and data quality.
Common-name guardrails
Built-in caps prevent highly common names from achieving critical risk scores without corroborating context.
Evidence Capsules
Every match provides transparent, per-field confidence contributions so reviewers can inspect the reasoning.
FAQ
Frequently Asked Questions
What is a false positive in sanctions screening?
A false positive occurs when the screening system flags an individual or entity as a potential sanctions match, but upon review, they are found to be a different person or entity with a similar name.
What is an acceptable false positive rate?
Industry averages range from 5% to 15%, but 'acceptable' depends entirely on your risk appetite, the quality of your customer data, and your regulatory environment. A rate over 10% usually indicates a need for system tuning or better data collection.
How do context fields reduce false positives?
Context fields like Date of Birth or Country allow screening algorithms to confidently discount a match. For example, if 'Juan Carlos' matches a sanctioned name, but the DOB is 30 years apart, an intelligent system can automatically lower the risk score.
What happens when a false positive is flagged?
The transaction or onboarding is typically paused while a compliance analyst manually reviews the alert. They inspect evidence, confirm the mismatch, mark it as a false positive in the audit log, and release the block.
Related products