Bias and Fairness

Unequal outcomes from AI in hiring, credit, benefits and pricing.

A University of Washington study of three LLMs ranking resumes preferred white-associated names 85% of the time and Black-associated names 9%. Enforcement is thin: New York State auditors found 17 potential violations of NYC's hiring-AI law where the city regulator found one.

Names preferred in LLM resume screening: White-associated 85%; Male-associated 52%; Female-associated 11%; Black-associated 9%. University of Washington, Oct 2024 NAMES PREFERRED IN LLM RESUME SCREENING White-associated 85% Male-associated 52% Female-associated 11% Black-associated 9% PERCENT OF COMPARISONS UNIVERSITY OF WASHINGTON, OCT 2024
iTutorGroup applicants auto-rejected
200+
UK 2020 grades lowered by algorithm
39.1%
NYC violations: state vs city
17 vs 1
Registry incidents
0

Evidence

Hiring software drew the first US class action; China banned algorithmic price discrimination

Documented cases: bias and Fairness
DateCaseCountryWhat happenedSource
Sep 11, 2023EEOC v. iTutorGroupUnited States$365,000; software rejected older applicantsSource
May 16, 2025Mobley v. WorkdayUnited StatesAge-bias claim certified as a nationwide collectiveSource
Dec 2, 2025NY State audit of Local Law 144United States32 companies reviewed; regulator found 1 likely violation, auditors 17Source
Dec 2024DWP Universal Credit fraud modelUnited KingdomDepartment found statistically significant disparities by age, disability and nationalitySource
Apr 10, 2026Internet Platform Pricing RulesChinaBan on algorithmic price discrimination by willingness to paySource

Registry

No registry incident turns on bias and fairness yet

The cases above are documented, but none yet meets the registry bar of a named developer, a failed control and a known finder.

Rules

Colorado softened its AI law; China and the EU tightened theirs

Rules and frameworks: bias and fairness
RuleJurisdictionWhat it requiresStatus
Colorado SB 26-189 (replaces SB 24-205)ColoradoNotice and explanation after adverse outcomesEffective Jan 1, 2027
NYC Local Law 144New York CityAnnual bias audit and candidate noticeIn force since Jul 2023
EU AI Act Article 10EUBias testing for high-risk AI training dataDec 2, 2027
Internet Platform Pricing RulesChinaNo algorithmic price discriminationEffective Apr 10, 2026

Status colors: green in force, blue voluntary standard or framework, amber adopted and phasing in, gray draft or pending, red withdrawn.

Record

Page Record

Datasets
ai_trust_evidence.csv (5 rows); registry incidents (0 linked)
Key sources
EEOC v. iTutorGroup
Mobley v. Workday
NY State audit of Local Law 144
DWP Universal Credit fraud model
Internet Platform Pricing Rules
Last verified
October 2026