Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This study evaluates nine pre-trained embedding models for AI-driven recruitment, focusing on their susceptibility to gender bias in applicant scoring. Using the synthetic FairCVdb dataset, the authors demonstrate that while explicit gender scrubbing reduces gender leakage, it does not fully eliminate it. Additionally, they implement a multi-task adversarial learning framework that enhances fairness in scoring but primarily benefits original biographies, highlighting the need for combined debiasing strategies.
Explicit gender scrubbing cuts gender leakage in recruitment models, but adversarial learning is essential for achieving true fairness.
AI-based recruitment systems that rely on machine learning models trained on historical CV data, risk perpetuating and amplifying social biases. A key challenge arises in unstructured CV text, where pre-trained language model embeddings may infer sensitive attributes such as gender even after explicit indicators are removed. In this paper, we evaluate nine pre-trained embedding models on the synthetic FairCVdb dataset, analyzing the informativeness of their embeddings for applicant scoring and their susceptibility to gender leakage, on both original and gender-scrubbed biographies. We further use a multi-task adversarial learning framework with gradient reversal to predict applicant suitability while suppressing gender information from learned representations. Finally, we use a multi-objective Pareto-front-based model selection to balance predictive utility and fairness. Our experimental results show that explicit gender scrubbing substantially reduces but does not eliminate gender leakage, while adversarial learning improves fairness mainly on original biographies and acts as a complementary strategy rather than a substitute for text-level debiasing.