IBI Score
8.0
out of 10
Severity4_grave
Frequencysystematique
Scope groups4
Detection difficulty4
🇫🇷 Définition (FR)
Dossiers médicaux électroniques contenant des biais historiques reproduits par les modèles.
🇬🇧 Definition (EN)
Electronic health records containing historical biases reproduced by models.
TriAxis Taxonomy
Structural Classification
O
Origin
O1
Systemic
L
Lifecycle
L1
Historical
S
Specific
S24
Institutional bias
💡
Concrete Example
Les notes médicales contiennent des termes stigmatisants pour les patients noirs, reproduits par les LLMs.
📊 Detection Metrics
Demographic Parity DifferenceDisparate Impact RatioEqualized Odds Difference
⚙ Alert Thresholds
Disparate Impact Ratio0.8
Equalized Odds Difference0.1
Demographic Parity Difference0.1
📚 Sources & Scientific References
- •Characterization of Stigmatizing Language in Medical RecordsKeith Harrigian, Ayah Zirikly, Brant Chee, Alya Ahmad, Anne Links, Somnath Saha, Mary Catherine Beach, Mark Dredze (2023)
- •
- •A Framework for Understanding Selection Bias in Real-World Healthcare DataRitoban Kundu, Xu Shi, Jean Morrison, Jessica Barrett, Bhramar Mukherjee (2024)
🛡 Impact & Mitigation
🎯 Vulnerable groups impacted
Black / Afro-descendant
🛠 Mitigation methodologies
- AUDIT_TIERS : Third-party Audit
🏢 Sectors concerned
Healthcare & Medicine (Risk: /5)
⚖ Applicable regulations
🇪🇺 Europe
- •AI_ACT_10 : AI Act (UE) 2024/1689
- •RGPD_9 : RGPD (UE) 2016/679
- •RGPD_22 : RGPD (UE) 2016/679
🌍 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
🌐 autre
- •IN_DPDP_RULES : Inde - Digital Personal Data Protection Rules
- •BR_LGPD : Brésil - LGPD
📋 Metadata
Contributor
Mohamadou SIDIBE
Institution
AI BIAIS INITIATIVE
Version
1
Created on
October 14, 2026
Updated on
November 4, 2026
UUID
e018f701-607e-4f3e-8e99-9fe19343093d
Explore other biases
250 documented biases, classified by the TriAxis taxonomy.