AI Self-preferencing in Algorithmic Hiring: Empirical Evidence and Insights (PDF)
Using a large-scale controlled resume correspondence experiment, we find that LLMs consistently prefer resumes generated by themselves over those written by humans or produced by alternative models, even when content quality is controlled. The bias against human-written resumes is particularly substantial, with self-preference bias ranging from 67% to 82% across major commercial and open-source models. To assess labor market impact, we simulate realistic hiring pipelines across 24 occupations. These simulations show that candidates using the same LLM as the evaluator are 23% to 60% more likely to be shortlisted than equally qualified applicants submitting human-written resumes, with the largest disadvantages observed in business-related fields such as sales and accounting. We further demonstrate that this bias can be reduced by more than 50% through simple interventions targeting LLMs' self-recognition capabilities.
Из интересного: предпочтение собственной генерации коррелирует со способностями LLM к распознанию собственного вывода, которая, в свою очередь, коррелирует с количеством весов. Описанные авторами "simple interventions" включают в себя два способа. Первый — явное включение в промпт указание на игнорирование авторства текста. Второй — использование ансамбля LLM, включающих в себя модели с более слабыми способностями к самораспознаванию.