Home›Biases›IBI-00181
IBI-001815 - Criticalpublie

LLM Credit Allocation Bias

Biais d'allocation de crédit par LLM

IBI Score
8.7
out of 10
Severity5_critique
Frequencyfrequent
Scope groups5
Detection difficulty4
🇫🇷 Définition (FR)

LLMs recommandant des niveaux de risque ou de crédit différents selon le profil perçu.

🇬🇧 Definition (EN)

LLMs recommending different risk or credit levels based on perceived profile.

TriAxis Taxonomy

Structural Classification

O
Origin
O2
Computational & Statistical
L
Lifecycle
L7
Deployment
S
Specific
S18
Group attribution bias
💡
Concrete Example

Un LLM refuse un prêt fictif à un profil « afro-américain » mais l'accepte pour un profil « blanc ».

📊 Detection Metrics

Distribution ShiftDemographic Parity Difference

⚙ Alert Thresholds

Distribution Shift0.1
Demographic Parity Difference0.1

📚 Sources & Scientific References

🛡 Impact & Mitigation

🎯 Vulnerable groups impacted

Black / Afro-descendantArab / MENA personsWorking classMinority language speakersLGBTQ+ personsWomenNon-binary personsRural areasSeniors (60+)Youth (18-25)Asian personsMotor disability

🛠 Mitigation methodologies

  • DEMO_PARITY : Demographic Parity
  • COUNTERFACT : Counterfactual Fairness
  • AUDIT_TIERS : Third-party Audit

🏢 Sectors concerned

Finance & Banking (Risk: /5)

⚖ Applicable regulations

🇪🇺 Europe
  • •
    AI_ACT_10 : AI Act (UE) 2024/1689
  • •
    RGPD_22 : RGPD (UE) 2016/679
🇺🇸 United States
  • •
    US_TX_TRAIGA : Texas Responsible AI Governance Act (HB 149)
🌍 International
  • •
    ISO_42001 : ISO/IEC 42001:2023
  • •
    UNESCO_2021 : UNESCO - Recommandation sur l'éthique de l'IA
  • •
    OECD_AI_PRINCIPLES : OCDE - Principes IA
  • •
    COE_AI_CONVENTION_INT : Conseil de l'Europe - Convention IA
🇫🇷 France
  • •
    FR_CNIL_AI : CNIL - Recommandations IA
🌐 Canada
  • •
    CA_AIDA : Canada - Loi sur l'IA et les données (AIDA)

📋 Metadata

Contributor
Mohamadou SIDIBE
Institution
AI BIAIS INITIATIVE
Version
1
Created on
October 14, 2026
Updated on
November 4, 2026
UUID
a3d4da53-863d-4c5c-8a5c-b70e2218daad

Explore other biases

250 documented biases, classified by the TriAxis taxonomy.