Disparate Accuracy in Biometric Recognition
生物特徵辨識準確度差異
Verification & Validation
Risk Description
When an identity verification system exhibits a markedly higher failure rate for a particular group, requiring those users to undergo additional manual checks every time; due to unmitigated control gaps, the organization evaluated system performance using overall accuracy without disaggregating by group, and remained unaware of the disparity for a long time, 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
Replace single aggregate metrics with cross-group accuracy testing, examining failure rates per group; Train on diverse datasets to strengthen recognition of under-served groups; Provide equivalent alternative verification for high-failure groups to avoid unequal burden