Evidence Chain & Source Traceability

Every match returned by SanctionsWise API includes a complete evidence chain, providing full audit trail for regulatory compliance and internal investigations.

---

Why Evidence Chain Matters

Regulatory Requirements

Financial institutions must demonstrate due diligence in their sanctions compliance programs:

• BSA/AML: Bank Secrecy Act requires documented screening processes
• OFAC Compliance: Must show source of sanctions information
• FINRA: Broker-dealers need auditable compliance records
• European Regulations: GDPR requires data provenance tracking

SanctionsWise Evidence Chain

Every match includes:

1. Source Authority - Who published the sanctions list
2. Source List ID - Which specific list the entity appears on
3. Source Entry UID - Original identifier from the authority
4. Data Timestamps - When data was retrieved and last updated
5. Match Details - How the match was determined

---

Evidence Chain Structure

Complete Match Response

{
  "screening_id": "SCR-20251210185252-2FFDC8AE",
  "status": "potential_match",
  "query": {
    "name": "Vladimir Putin",
    "entity_type": "individual",
    "threshold": 0.60,
    "lists_screened": ["ofac_sdn"]
  },
  "matches": [
    {
      "entity_id": "sw_ofac_sdn_35096",
      "matched_name": "VLADIMIR VLADIMIROVICH PUTIN",
      "confidence_score": 0.7714,
      "match_type": "fuzzy+semantic",
      "semantic_score": 0.89,

      "entity_details": {
        "primary_name": "VLADIMIR VLADIMIROVICH PUTIN",
        "entity_type": "individual",
        "nationality": "Russia",
        "programs": ["RUSSIA-EO14024", "UKRAINE-EO13660"],
        "listing_date": "2022-02-25"
      },

      "source_evidence": {
        "source_authority": "US Treasury OFAC",
        "source_list_id": "ofac_sdn",
        "source_entry_uid": "sw_ofac_sdn_35096",
        "last_updated": "2025-12-10T17:26:38Z"
      },

      "identifier_evidence": {
        "match_type": "exact",
        "query_type": "passport",
        "entity_type": "passport_number",
        "confidence": 1.0,
        "matched_value": "AB12***"
      }
    }
  ],
  "timestamp": "2025-12-10T18:52:52.527387Z",
  "processing_time_ms": 378,
  "data_freshness": {
    "last_update": "2025-12-10T17:26:38Z"
  }
}

---

Evidence Components

1. Screening Identification

Field	Description	Example
screening_id	Unique request identifier	SCR-20251210185252-2FFDC8AE
timestamp	When screening was performed	2025-12-10T18:52:52.527387Z

Purpose: Links the screening request to audit logs and compliance records.

2. Query Context

"query": {
  "name": "Vladimir Putin",
  "entity_type": "individual",
  "threshold": 0.60,
  "lists_screened": ["ofac_sdn"]
}

Field	Description
name	Original query submitted
entity_type	Classification used for matching
threshold	Confidence cutoff applied
lists_screened	Which sanctions lists were checked

Purpose: Documents exactly what was searched and how.

3. Source Authority

"source_evidence": {
  "source_authority": "US Treasury OFAC",
  "source_list_id": "ofac_sdn",
  "source_entry_uid": "sw_ofac_sdn_35096",
  "last_updated": "2025-12-10T17:26:38Z"
}

Field	Description
source_authority	Government body that issued the sanctions
source_list_id	Specific list within that authority
source_entry_uid	Original entry identifier
last_updated	When entity data was last modified

Purpose: Traces match back to authoritative government source.

4. Match Methodology

{
  "confidence_score": 0.7714,
  "match_type": "fuzzy+semantic",
  "semantic_score": 0.89
}

Field	Description
confidence_score	Combined confidence (0.0-1.0)
match_type	How match was determined
semantic_score	AI similarity score (if used)

Match Types:

• exact - Direct name match after normalization
• fuzzy - String similarity algorithms
• fuzzy+semantic - Enhanced with AI vectors
• semantic - Found via contextual search
• identifier_confirmed - Document number match

Purpose: Explains why this was flagged as a match.

5. Data Freshness

"data_freshness": {
  "last_update": "2025-12-10T17:26:38Z"
}

Purpose: Proves sanctions data was current at time of screening.

---

Source Authorities

Supported Authorities

Authority	Lists	Update Frequency
US Treasury OFAC	SDN, Consolidated, FSE, SSI	Real-time to daily
US Commerce BIS	Entity List, Denied Persons	Weekly

Authority Verification

Each entity's source can be verified against original government publications:

List	Verification URL
OFAC SDN	https://sanctionslist.ofac.treas.gov/
OFAC Consolidated	https://www.treasury.gov/ofac/downloads/
BIS Entity List	https://www.bis.doc.gov/index.php/the-denied-persons-list

---

Audit Trail Best Practices

1. Store Complete Responses

Always store the full API response for compliance records:

import json
from datetime import datetime

def screen_and_log(name, entity_type):
    result = screen_entity(name, entity_type)

    # Store complete audit record
    audit_record = {
        "screening_id": result["screening_id"],
        "request_time": datetime.utcnow().isoformat(),
        "query": result["query"],
        "status": result["status"],
        "match_count": result["match_count"],
        "matches": result["matches"],
        "data_freshness": result["data_freshness"]
    }

    save_to_audit_database(audit_record)
    return result

2. Log Decision Points

Document when matches require escalation:

def process_screening_result(result):
    if result["status"] == "clear":
        log_decision("APPROVED", result["screening_id"], "No matches")

    elif result["status"] == "match":
        log_decision("BLOCKED", result["screening_id"],
                    f"High-confidence match: {result['matches'][0]['matched_name']}")
        escalate_to_compliance(result)

    elif result["status"] == "potential_match":
        log_decision("REVIEW_REQUIRED", result["screening_id"],
                    f"{result['match_count']} potential matches")
        queue_for_manual_review(result)

3. Maintain Evidence Retention

Keep screening records for required retention periods:

Regulation	Retention Period
BSA/AML	5 years
OFAC	5 years
GDPR	As long as necessary

4. Regular Data Freshness Checks

Verify screening was performed against current data:

from datetime import datetime, timedelta

def validate_data_freshness(result):
    last_update = datetime.fromisoformat(
        result["data_freshness"]["last_update"].replace("Z", "+00:00")
    )

    age_hours = (datetime.utcnow() - last_update.replace(tzinfo=None)).seconds / 3600

    if age_hours > 24:
        log_warning(f"Screening used data {age_hours:.1f} hours old")

    return age_hours < 24

---

Compliance Reporting

Generate Audit Report

def generate_compliance_report(screenings):
    """Generate compliance audit report from screening records."""

    report = {
        "report_date": datetime.utcnow().isoformat(),
        "total_screenings": len(screenings),
        "status_breakdown": {
            "clear": sum(1 for s in screenings if s["status"] == "clear"),
            "match": sum(1 for s in screenings if s["status"] == "match"),
            "potential_match": sum(1 for s in screenings if s["status"] == "potential_match")
        },
        "sources_used": list(set(
            s["query"]["lists_screened"][0]
            for s in screenings
        )),
        "average_confidence": sum(
            s["matches"][0]["confidence_score"]
            for s in screenings if s["matches"]
        ) / max(1, sum(1 for s in screenings if s["matches"]))
    }

    return report

Sample Audit Report

SANCTIONS SCREENING AUDIT REPORT
================================
Report Date: 2025-12-10T19:00:00Z
Reporting Period: 2025-12-01 to 2025-12-10

SCREENING SUMMARY
-----------------
Total Screenings: 10,542
  - Clear: 10,012 (95.0%)
  - Potential Match: 518 (4.9%)
  - Confirmed Match: 12 (0.1%)

DATA SOURCES
------------
  - OFAC SDN: 10,542 screenings
  - Data freshness: Real-time (< 1 hour)

MATCH ANALYSIS
--------------
  - Average confidence (matches): 0.847
  - Match types: fuzzy (65%), fuzzy+semantic (28%), exact (7%)

EVIDENCE CHAIN INTEGRITY
------------------------
  - All screenings include source_evidence: YES
  - All matches traceable to authority: YES
  - Data freshness within 24 hours: YES

---

For API integration details, see the API Reference