研究揭示开源大模型在招聘中存在性别和种族偏见触发机制
Linguistic Triggers of Gender and Racial Bias in Open-Weight LLMs Applied to Recruitment
朋友，这篇论文挺有意思的，它不是简单说大模型有偏见，而是具体分析了哪些语言触发器（比如代理性语言、排斥性编码）会导致偏见，还给出了一个具体的审计方法，对做招聘AI系统的人应该挺有参考价值。
研究对六款开源大模型（如 Llama 3.2、Mistral）进行系统性审计，发现招聘启事语言会影响招聘结果。使用代理性语言会降低对女性候选人的推荐分数（相关系数 r_rb = 0.309），而使用共同体语言可部分抵消这一影响。同时，带有排斥性编码的语言会显著降低非白人招聘者的评分（r_rb = 0.646-0.758），并抑制非白人求职者表达兴趣。研究还提出了一套预部署审计协议，用于评估和规避这些偏见。
Linguistic Triggers of Gender and Racial Bias in Open-Weight LLMs Applied to Recruitment
Open-weight large language models are rapidly entering hiring pipelines, yet their discriminatory failure modes -- and the regulatory exposure these create under the EU AI Act high-risk classification (Annex III) and U.S. EEOC adverse-impact analysis -- remain poorly understood. We present the first systematic, multi-model audit of open-weight LLMs that treats job-posting language as the primary experimental variable, evaluating six models (Llama 3.2, Mistral, Gemma 3, Qwen 3, Phi 3, DeepSeek-R1) across four controlled experiments that jointly probe recruiter-simulation and job-seeker-simulation tasks. We find that (1) agentic posting language depresses recruiter recommendation scores for female candidates (r_rb = 0.309, p_Bonf = 7x10^-5; model-fixed-effects r_rb = 0.448), while communal language partially reverses the penalty; and (2) coded-exclusion language suppresses non-White recruiter scores at large effect sizes (r_rb = 0.646-0.758) and, on the job-seeker side, selectively deters non-White personas from expressing interest -- operationalizing a chilling-effect mechanism at scale. A label-ablation experiment isolates the explicit demographic persona label as the primary causal driver, and Word Embedding Association Tests corroborate these findings at the representational level (d = 1.01-1.45 under Caliskan et al.'s multi-word gender attribute lists). We translate these results into a concrete pre-deployment audit protocol -- posting-vocabulary scoring, persona-conditioned LLM probing, and adverse-impact flagging against the four-fifths threshold -- that operationalizes the documentation and risk-management obligations Annex III imposes on high-risk AI in recruitment.
