RESEARCHInvestigateNEXT 12 MONTHS
Linguistic Triggers of Gender and Racial Bias in Open-Weight LLMs Applied to Recruitment
arXiv cs.CL — Computation and Language
Factual evidence
What the source reports
Research shows open-weight LLMs in recruitment display bias triggered by job posting language, impacting EU AI Act and EEOC compliance.
OneBench interpretation
Institutional assessment
So what
Open-weight models used in recruitment present compliance risks under high-risk EU AI Act classifications and EEOC discrimination rules.
Do what
Review bias-testing frameworks for automated recruitment systems with the team responsible for model risk management.