logo
welcome
Ars Technica

Ars Technica

AIs show distinct bias against Black and female résumés in new study

Ars Technica
Summary
Nutrition label

90% Informative

University of Washington researchers ran hundreds of publicly available résumés and job descriptions through three different Massive Text Embedding (MTE) models.

They used the models to generate embedded relevance scores for each résumé and job description pairing.

In all three models, white names were preferred in all three MTE models, compared to Black names being preferred in just 8.6 percent of tests.

VR Score

96

Informative language

99

Neutral language

51

Article tone

formal

Language

English

Language complexity

66

Offensive language

not offensive

Hate speech

not hateful

Attention-grabbing headline

not detected

Known propaganda techniques

not detected

Time-value

long-living

Affiliate links

no affiliate links