Sanctions Screening Alert Disposition

The Problem

Banks screen all customers and transactions against OFAC, EU, UN, and other sanctions lists. Screening matches generate alerts requiring investigation and disposition. Alert volume is enormous: 10,000-50,000+ alerts/month at large banks. However, 85-95% of hits are false positives: name-only matches with insufficient identifying information (e.g., "John Smith" matches to a sanctioned entity but customer is a different John Smith with different DOB/location). Current process: compliance team manually reviews each alert, gathers customer identifying data, determines if match is true positive or false positive, and documents disposition. Median time: 10-20 minutes per alert for obvious false positives; 30-60 minutes per alert for more ambiguous cases. Backlog of unreviewed alerts creates regulatory risk.

Data Requirements

Data Sources:

• Sanctions Screening System (Actimize by NICE, Fircosoft, SAS AML, or custom) , Screening alerts, match details, match scores, screening run history
• OFAC Sanctions List API , U.S. Treasury Office of Foreign Assets Control consolidated list
• EU Sanctions Database API , European Union consolidated sanctions list, member state lists
• UN Consolidated Sanctions List , UNSC sanctions screening data
• Customer KYC Database (CIP, KYC system) , Customer full legal name, DOB, address, nationality, business type, beneficial owner data
• Customer Transaction Database , Historical transactions, counterparties, geographic corridors, transaction patterns
• Beneficial Owner Registry (internal) , Ownership structure, ultimate beneficial owners, related parties
• Disposition Rules Database (internal) , False positive thresholds, disposition logic, name matching rules, distinguishing data requirements
• AML Case Management System , Prior sanctions dispositions, similar alerts, investigator decisions

Data Classification:

• Public: Sanctions lists (OFAC, EU, UN)
• Internal Structured: Screening alerts, customer KYC data, transaction history, disposition rules, match scores
• Internal Unstructured: Investigator notes, disposition narratives, prior case files
• External Third-Party: OFAC/EU/UN sanctions feeds
• Sensitive/Regulated: Customer personal data (DOB, address), KYC information, beneficial owner data, transaction history, investigation records

Data Quality Requirements:

Sanctions list freshness: daily updates from OFAC, EU sources. Customer data completeness: 99% (full name, DOB, address available). Customer data accuracy: 100% (must match source records). Transaction history completeness: 100% for lookback period (typically 3-5 years). Match score accuracy: standardized per screening system (e.g., Actimize match scoring). Disposition accuracy baseline: 90%+ for obvious false positives (validated from historical investigator decisions).

Integration Complexity: High , Requires integration with screening system (Actimize, Fircosoft, custom) with real-time alert feeds. Sanctions list integration with daily updates from OFAC, EU, UN sources (different formats, different refresh cycles). Customer data enrichment from KYC system, transaction database, beneficial owner registry. Investigator workflow integration with AML case management system. Disposition rules require complex business logic and regular updates as sanctions policy changes. High regulatory scrutiny requires extensive audit logging and monitoring.

Regulatory Alignment

• OFAC Regulations (31 CFR Chapter V): Agent must document all screening and disposition; false negatives are compliance violations
• EU Sanctions Regulation: Agent must comply with EU member state sanctions screening procedures
• FinCEN Guidance on Sanctions Compliance: Agent assists in disposition; human specialist retains authority for true positive determination
• AML/CFT (FATF, Basel Committee): Sanctions screening is mandatory control; agent supports efficiency but not decision authority

Comparable Implementations

• Commonwealth Bank: Sanctions alert disposition AI; 80% false-positive filtering; investigator efficiency improved by 60%
• Barclays: Sanctions screening alert triage; 75% of obvious false positives auto-cleared; investigation backlog reduced by 50%