BiasLab: A Multilingual Dual-Framing Framework for LLM Bias Measurement, Applied to Workplace and HR Contexts
Not on the map. Why?
Abstract (arXiv)
Background: Large language models (LLMs) harbor systematic biases that are particularly consequential in workplace and HR contexts, where their outputs increasingly influence hiring, job design, and organizational decisions. Existing bias-evaluation approaches remain methodologically fragmented, limiting practitioners' ability to assess deployment risks. Objective: This study introduces BiasLab, a multilingual dual-framing framework to quantify and compare directional output-level bias in LLMs, demonstrated across six workplace and HR-relevant topics. Methods: BiasLab combines mirrored affirmative and reverse prompt pairs, randomized wrapper perturbations, fixed-choice response constraints, and polarity-aligned scoring. Ten LLMs were evaluated across six topics (gender in leadership, employment gap candidates, age in hiring, remote versus office work, four-day versus five-day work weeks, and AI-assisted versus human-only hiring), spanning 12 languages and 30 iterations per framing direction, yielding 43,200 responses. Results: All ten models showed consistent directional preferences across every topic. A recurring asymmetric pattern emerged in which models rejected disfavored claims more strongly than they endorsed their opposites, a distinction invisible to single-frame designs. Conclusions: BiasLab provides a standardized, reproducible instrument for measuring directional preferences across models. Whether a preference constitutes bias in a fairness sense is topic-dependent: for protected attributes such as gender and age it maps onto equal-employment standards, whereas elsewhere it is better described as systematic preference. The framework lets organizations compare and vet models before adopting them for hiring.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 20 of its 34 references that are in the map. First posted on arXiv 11 Jan 2026.
Builds on
20 of 34 references.
References · 20
- 2023ChatGPT Outperforms Crowd-Workers for Text-Annotation Taskscited by 870
- 2023Bias and Fairness in Large Language Models: A Surveycited by 639
- 2021Recent Advances in Natural Language Processing via Large Pre-Trained Language Models: A Surveycited by 463
- 2023Large Language Models in Finance: A Surveycited by 226
- 2023Large language models in healthcare and medical domain: A reviewcited by 127
- 2023Generative Language Models Exhibit Social Identity Biasescited by 111
- 2024The Political Preferences of LLMscited by 84
- 2024Bias in Large Language Models: Origin, Evaluation, and Mitigationcited by 77
- 2022Gender Bias in Masked Language Models for Multiple Languagescited by 60
- 2023Large Language Models in Education: Vision and Opportunitiescited by 54
- 2024Gender, Race, and Intersectional Bias in Resume Screening via Language Model Retrievalcited by 51
- 2024LLM-GLOBE: A Benchmark Evaluating the Cultural Values Embedded in LLM Outputcited by 27
8 more
- 2024Large Language Models are Biased Because They Are Large Language Modelscited by 26
- 2024JobFair: A Framework for Benchmarking Gender Hiring Bias in Large Language Modelscited by 21
- 2024Understanding Intrinsic Socioeconomic Biases in Large Language Modelscited by 9
- 2025No Thoughts Just AI: Biased LLM Hiring Recommendations Alter Human Decision Making and Limit Human Autonomycited by 8
- 2025Mapping Geopolitical Bias in 11 Large Language Models: A Bilingual, Dual-Framing Analysis of U.S.-China Tensionscited by 7
- 2025More or Less Wrong: A Benchmark for Directional Bias in LLM Comparative Reasoningcited by 7
- 2025Gender and Positional Biases in LLM-Based Hiring Decisions: Evidence from Comparative CV/R'esum'e Evaluationscited by 4
- 2025CEA-LIST at CheckThat! 2025: Evaluating LLMs as Detectors of Bias and Opinion in Textcited by 1
Built on by
Not cited yet.