LLM bias in salary recommendations across gender, age, and nationality
Published: 26 May 2026| Version 1 | DOI: 10.17632/33stgn5vcp.1
Contributor:
Tanapong PotipitiDescription
To examine bias in LLM salary recommendations, we generate synthetic candidate and job profiles and systematically match each candidate with each job. The LLMs, framed as HR compensation analysts, are asked to propose a competitive annual salary for each job-candidate pairing.
Files
Institutions
- Chulalongkorn UniversityBangkok, Bangkok
Categories
Labor Economics, Labor Discrimination, Large Language Model