# Small Changes, Big Impact: Demographic Bias in LLM-Based Hiring Through Subtle Sociocultural Markers in Anonymised Resumes

arXiv: https://arxiv.org/abs/2603.05189 · HTML: https://mlsanity.com/paper/2603.05189/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 5 Mar 2026.

## Built on by (most cited first, 4 of 4)

- [Gender Bias in LLM Hiring Decisions: Evidence from a Japanese Context and Evaluation of Mitigation Strategies](https://mlsanity.com/paper/2606.18649/) (2026)
- [Beyond the Name: Demographic Leakage in De-Identified R'esum'es and Evaluation Artifacts in LLM Bias Audits](https://mlsanity.com/paper/2609.16501/) (2026)
- [Fairness Beyond Anonymization? Demographic Leakage in German LLM-Generated Resumes](https://mlsanity.com/paper/2609.22188/) (2026)
- [Homogenization in Multi-Agent Systems](https://mlsanity.com/paper/2610.09824/) (2026)

## Abstract (arXiv)

Large Language Models (LLMs) are increasingly deployed in resume screening pipelines. Although explicit PII (e.g., names) is commonly redacted, resumes typically retain subtle sociocultural markers (languages, co-curricular activities, volunteering, hobbies) that can act as demographic proxies. We introduce a generalisable stress-test framework for hiring fairness instantiated in the Singapore context: 100 neutral job-aligned resumes are augmented into 4100 variants spanning four ethnicities and two genders, differing only in job-irrelevant markers. We evaluate 18 LLMs in two settings: (i) Direct Comparison (1v1) and (ii) Score & Shortlist (Top-Score Rates), each with and without rationale prompting. We find that even without explicit identifiers, models recover demographic attributes with high F1 and exhibit systematic disparities, with models favouring markers associated with Chinese and Caucasian males. Ablations show language markers suffice for inferring ethnicity, while hobbies and activities are utilised for gender. Furthermore, prompting for explanations may paradoxically amplify bias. Our findings suggest that seemingly innocuous markers surviving anonymisation can materially skew automated hiring outcomes.