S0079RS4-T16-S0079-Z · Full risk code

Group Bias in Credit Scoring

信用評分族群偏誤

Verification & Validation
Risk Description

When applicants with comparable financial profiles receive markedly different limits depending on the group to which they belong; due to unmitigated control gaps, only after affected individuals raise objections does the organization examine the model and find it reproducing historical differential treatment through indirect variables, triggering compliance exposure and operational reputational costs.

Framework Mappings

EU AI ActArt.10
NIST AI 600-1Harmful Bias and Homogenization
NIST AI RMFMEASURE 2.11
ISO/IEC TR 24027
ISO/IEC 5338驗證與確效
MIT AI Risk RepositoryDomain 1

Risk Treatment & Implementation Guidance

Monitor cross-group fairness in credit decisions—limit distributions and approval-rate gaps; Audit for bias, examining whether indirect variables reproduce historical discrimination; Apply debiasing to confirmed sources and validate the correction