ComplianceGRID
Risk Score Interpreter — Quick Reference Cheat Sheet · v1.0 · Feb 2026
BRS Business Risk Score | IRS Individual Risk Score
Scale: 0 – 1000 | AI Engine: Custom-trained Advanced Models
All data: publicly available sources only
Scale: 0 – 1000 | AI Engine: Custom-trained Advanced Models
All data: publicly available sources only
Score Bands — At a Glance
Critical
851–1000
Active sanctions match, federal criminal conviction, or multiple severe confirmed findings. Transacting may be illegal.
▶ Block immediately. Escalate to legal. File SAR if applicable.
High
601–850
Significant confirmed risk factors. High probability of regulatory concern. Enhanced due diligence required.
▶ Do not onboard without senior compliance approval. Request entity explanation.
Elevated
401–600
Multiple moderate factors or one significant concern with some mitigation evidence. Caution warranted.
▶ Enhanced due diligence. Require certifications. Monitor quarterly.
Moderate
201–400
Some risk signals present — minor violations, historical events, or low-confidence adverse signals only.
▶ Standard due diligence. Document review. Re-score annually.
Low
0–200
Minimal or no adverse public signals found. Standard onboarding appropriate. Note: reflects public data only.
▶ Proceed with standard onboarding. Re-score annually.
What the Same Score Means — By Use Case
| Score | AML / KYC / Sanctions | Carrier Onboarding | Vendor Due Diligence | Beneficial Owner |
|---|---|---|---|---|
| 750–1000 | Block. Potential prohibited transaction. Escalate to BSA officer. SAR assessment required. | Do not tender. OFAC or cargo fraud exposure unacceptable. Report to compliance team. | Reject vendor. Risk of regulatory blowback from association. Document decision. | Block account opening. Cannot certify BO. Escalate immediately. |
| 601–750 | Enhanced monitoring. Increased transaction review. Quarterly compliance file review. | Hold. Request safety docs & insurance. Do not tender until resolved. | Escalate to legal. Require indemnification clause and compliance certification. | Enhanced CDD. Additional ID verification + source-of-funds documentation. |
| 401–600 | Enhanced transaction monitoring. Semi-annual review. Document rationale for proceeding. | Proceed with additional insurance requirement and increased check-in cadence. | Require additional compliance reps & warranties in contract. Legal sign-off. | Proceed with enhanced CDD. Request explanation of flagged factors in writing. |
| 201–400 | Standard KYC sufficient. Annual review. Note file. | Onboard with standard agreement. Annual re-screen. | Approve with standard contract terms. No additional due diligence. | Certify BO. Proceed with account opening. Standard monitoring. |
| 0–200 | Standard KYC. Annual review cycle. | Onboard. Annual re-screen. | Approve. Standard terms. | Certify. Proceed. |
Special Response Flags — Always Check First
accusation_only: true
Factor is based on allegation (adverse media, indictment, social media) — NOT a confirmed finding. Weighted at 40% of base value. Score explanation will explicitly state this. Treat as a signal to investigate, not a conclusion.
confidence < 0.60
Low data coverage. Fewer public signals found than expected. Score should be treated as preliminary. Consider requesting a deep depth rescore or supplementing with manual review.
status: HISTORICAL_OLD
Factor event is >3 years old. Score contribution is discounted to 25% of base. Still flagged for transparency — evaluate whether remediation has occurred.
GDPR_NOTICE present (IRS only)
Subject is identified as an EU/EEA person. Score uses factual public record data only. Data subject rights notice included in response payload.
pattern_multiplier: 1.3
3+ factors in the same category detected. Score for that category is amplified 30%. Indicates a systemic pattern, not isolated incidents — higher concern than the raw factor count suggests.
How the Score Is Calculated
// 1. Each factor evaluated
CONFIRMED → 100% of base pts
PARTIAL → 60% of base pts
HISTORICAL ≤3y → 50% of base pts
HISTORICAL >3y → 25% of base pts
ALLEGATION → 40% of base pts
NOT_PRESENT → 0 pts
// 2. Industry weight applied
factor_score × industry_multiplier
// (0.2x–3.0x depending on vertical)
// 3. Pattern bonus
3+ factors in same cat → ×1.3
// 4. Mitigation credit
Documented remediation → −15%
// 5. Normalize to 0–1000
Max "clean" score → 850
// (850 cap reflects data limits)
CONFIRMED → 100% of base pts
PARTIAL → 60% of base pts
HISTORICAL ≤3y → 50% of base pts
HISTORICAL >3y → 25% of base pts
ALLEGATION → 40% of base pts
NOT_PRESENT → 0 pts
// 2. Industry weight applied
factor_score × industry_multiplier
// (0.2x–3.0x depending on vertical)
// 3. Pattern bonus
3+ factors in same cat → ×1.3
// 4. Mitigation credit
Documented remediation → −15%
// 5. Normalize to 0–1000
Max "clean" score → 850
// (850 cap reflects data limits)
Confidence Score Interpretation
0.90–1.0
High coverage. Score is reliable.
0.75–0.89
Good coverage. Minor gaps in some sources.
0.60–0.74
Moderate. Consider manual supplement.
0.40–0.59
Low. Request
depth: "deep" rescore.< 0.40
Very low. Entity may be obscured or new. Treat as preliminary only.
Low confidence does not mean low risk. It means less public data was found. A low-confidence CRITICAL score is still CRITICAL.
Highest-Weight Factors — By Category
Sanctions & Watchlists
Direct OFAC SDN match200 pts
BIS Entity List match180 pts
UN / EU / UK sanctions160–170 pts
OFAC exposure via officer/owner150–160 pts
PEP — current (IRS only)90 pts
Regulatory & Enforcement
SAM.gov debarment / suspension170–180 pts
HHS OIG LEIE exclusion160 pts
FINRA bar / suspension170 pts
SEC AAER enforcement action150–160 pts
FDA warning letter / import alert120 pts
Court & Legal
Federal criminal conviction190 pts
Federal criminal indictment130 pts ⚑
State felony conviction150 pts
Financial crime conviction180 pts
Chapter 7 / 11 bankruptcy70–100 pts
Reputational & Media
Adverse media: bribery / corruption130 pts ⚑
Adverse media: fraud allegations100–110 pts ⚑
Social: association w/ sanctioned entities80 pts ⚑
Social: extremist affiliation signals90 pts ⚑
Glassdoor repeated ethics complaints50 pts ⚑
⚑ = accusation_only flag likely — distinguish from confirmed findings in explanation field.
Industry Weight Highlights
Financial ServicesOFAC 3x · FINRA 3x · fraud 3x
Freight & LogisticsFMCSA 3x · OFAC 2.5x · DOT 2x
Healthcare & PharmaOIG LEIE 3x · DEA 3x · FDA 2.5x
Gov ContractorSAM debarment 3x · FCA 3x
Import / ExportBIS 3x · OFAC 3x · ITAR 3x
Firearms & DefenseATF 3x · OFAC 3x · criminal 3x
Controlled SubstancesDEA 3x · drug conviction 3x
AviationFAA cert 3x · OFAC 2.5x
Real Estate / PEBO sanctions 3x · shell co 2.5x
Technology / SaaSBIS/export 2x · sanctions 2x
Food & BeverageFDA alerts 3x · USDA 2.5x
Unknown IndustryOFAC always 2.5x min
Industry detected from declared NAICS/SIC code or inferred by AI from business name + SEC filings + web presence.
Social Media Intelligence — What Gets Flagged
All social signals are public content only. No authentication bypass. accusation_only: true always set on social-sourced factors. Two-stage pipeline: keyword pass → AI NLP for ambiguous signals.
5-Step Score Review Checklist
1
Check score band first
CRITICAL or HIGH? → stop, escalate before reading further. MODERATE or LOW? → continue standard process.
2
Check confidence score
Below 0.60? → treat as preliminary, consider deep rescore. Low confidence + HIGH band = still HIGH.
3
Scan accusation_only flags
Any factors flagged accusation_only: true? → note as unconfirmed. Read factor explanation for source and date before drawing conclusions.
4
Read top_factors in explanation
Top 3–5 factors are listed with source URL and date. Verify source directly for CRITICAL/HIGH scores before taking action.
5
Apply use-case action
Use the Use Case table (above) to determine the right action for your specific workflow. Carrier onboarding ≠ AML screening.
Quick Lookup — Key API Response Fields
score | Integer 0–1000. The composite risk score. |
band | LOW / MEDIUM / HIGH / CRITICAL |
bandExplanation | Plain-English band meaning with label, description, actionRequired, and nextSteps[]. |
summary.executiveSummary | Plain-English summary paragraph from AI. |
summary.recommendedActions | Array of action items based on risk findings. |
factors[].accusationOnly | true = allegation, not confirmed finding. |
factors[].matchConfidence | HIGH / MEDIUM / LOW — entity disambiguation confidence for this match. |
entityResolution.confidence | HIGH / MEDIUM / LOW — overall entity identification quality. |
entityResolution.inputCompleteness | 0–100% metric. Shows how complete your input is relative to recommended fields. |
jurisdictionRisk | Country risk level: SANCTIONED / HIGH / MEDIUM. Programs[] and notes when applicable. |
meta.modelVersion | Scoring model version. Track for score stability across re-scores. |