AI evaluators favor job applications written by certain AI models, independent of candidate quality
4.40T1 sourcearXiv cs.MA
arXiv cs.MAoriginal source ↗
Source record
Published by arXiv cs.MA (T1 source). The original is at https://arxiv.org/abs/2610.41866.
Pipeline notes
The summary and note below are generated by the signal pipeline — they are Beyond Desk’s reading, not quotations from the source.
SummaryA multi-model audit pairs eleven LLM writing agents with nine evaluator agents across ten job descriptions and fifty candidates per job. It reports model-dependent acceptance, local-model disadvantage, evaluator self-preference, and candidate filtering caused by model mismatch, linked to stylistic cues.
Why it mattersProvides a reproducible audit design and quantified evidence for teams assessing model choice, evaluator bias, and screening outcomes in AI-mediated hiring workflows.
Cited by
No citations on record.