Skip to main content

False Positive Management

Last Updated: 2025-12-28 Status: Complete

This page covers strategies for managing and reducing false positives in sanctions screening.

The False Positive Problem​

Industry Benchmark: 90-95%+ of sanctions screening alerts are false positives (not actual matches).

Historical vs. Current View​

Historical View: High false positive rates were once viewed as "better safe than sorry."

Current View (2025): High false positive rates indicate poor technology and poor data quality, not cautious compliance. Regulators and auditors now expect:

  • Effective screening systems
  • Tuned algorithms
  • Reduced manual review burden
  • Evidence of continuous improvement

Why False Positives Occur​

  1. Common names: "John Smith" matches thousands of people
  2. Over-sensitive thresholds: Catching too many low-probability matches
  3. Insufficient data points: Screening name only without DOB/address
  4. Poor data quality: Typos in merchant application data
  5. Lack of secondary screening: Not using entity resolution

Mitigation Strategies​

1. Use Secondary Identifying Data​

When screening, include additional data points beyond just names:

  • Date of birth
  • Address/location
  • Nationality/citizenship
  • Passport/ID numbers (when available)
  • Business registration numbers

Example:

  • Name match: "Ahmed Khan" → 500 potential matches
  • Name + DOB match: "Ahmed Khan, 1975-03-15" → 3 potential matches
  • Name + DOB + Country: "Ahmed Khan, 1975-03-15, Pakistan" → 1 match

2. Apply Risk-Based Thresholds​

  • Higher thresholds for common names
  • Lower thresholds for uncommon names
  • Adjusted thresholds by geography
  • Industry-specific tuning
Name CommonalityRecommended Threshold
Very common (top 100 names)95%+
Common90%+
Uncommon85%+
Rare80%+

3. Implement Entity Resolution​

Use AI/ML to:

  • Link related records
  • Identify unique entities
  • Consolidate duplicate alerts
  • Learn from historical review decisions

Entity Resolution Benefits:

  • Reduces duplicate alerts for same person
  • Builds confidence scores over time
  • Identifies connected parties

4. Continuous Tuning​

  • Track false positive rates by name type
  • Analyze review outcomes
  • Adjust thresholds quarterly
  • Update exclusion rules
  • Improve data collection quality

5. Maintain Exclusion Lists​

  • Document recurring false positives
  • Create "known good" lists
  • Require periodic review (annually)
  • Justify exclusions with evidence

Exclusion List Management​

Exclusion List Risks

While exclusion lists reduce false positives, they create compliance risk. A person initially excluded as false positive could later be added to SDN list. Regular re-screening of exclusions is essential.

Best Practices for Exclusion Lists​

Creating Exclusions:

  1. Document why the match is false positive
  2. Record specific identifying data that differentiates
  3. Require manager approval for each exclusion
  4. Set expiration date (12 months maximum)

Maintaining Exclusions:

  1. Re-screen exclusion list monthly against SDN updates
  2. Require annual re-validation of each exclusion
  3. Remove exclusions immediately if new information emerges
  4. Audit exclusion decisions quarterly

Documentation Required:

  • Original alert details
  • Investigation steps taken
  • Differentiating information
  • Approver name and date
  • Expiration date
  • Re-validation history

Measuring False Positive Rates​

Key Metrics​

MetricFormulaTarget
False Positive RateFalse Positives / Total Alerts<90%
True Positive RateTrue Positives / Actual Positives>99%
Review TimeAverage time to close alert<15 minutes
Escalation RateEscalations / Total Alerts<5%

Improvement Tracking​

  • Track metrics monthly
  • Compare against industry benchmarks
  • Set quarterly improvement targets
  • Report to compliance committee

Technology Solutions​

AI/ML-Powered Screening​

Modern solutions use machine learning to:

  • Learn from historical decisions
  • Automatically classify low-risk alerts
  • Prioritize high-risk matches
  • Suggest resolution actions

Workflow Automation​

  • Auto-close obvious false positives
  • Route complex cases to specialists
  • Track SLAs and aging
  • Generate audit reports
Share: