Counterfactual Bias Testing for Application Tracking System

A new arXiv paper (2608.26899v1) introduces a reusable LLM-agent-based methodology for auditing candidate-job matching systems for demographic bias, generating a K x (1+N) correspondence-audit matrix across five protected-characteristic axes and computing a nine-metric fairness suite. In tests on 5 job orders, 100 base candidates, and 10 demographic treatments, score shifts, top-K retention, and merit-aware rate gaps stayed within tolerance, but rank-stability (MARC) and nDCG@K surfaced borderline findings, including on the neutral baseline, arguing for multi-metric auditing over single aggregate scores.

arXiv:2608.26899v1 Announce Type: new Abstract: Automated candidate-job matching systems are increasingly classified as high-risk AI under emerging regulation, yet auditing them for demographic bias is expensive: classical correspondence-audit studies require hand-crafted resumes and manual submission, which does not scale to fast pipeline retraining cycles. This paper presents a general, reusable methodology that 1 uses task-specialized LLM agents to synthesize identity-neutral base resumes and inject controlled demographic treatments across five protected-characteristic axes sex/gender, age, residence, language, disability , producing a K x 1+N correspondence-audit matrix; 2 qualitatively flags inferred protected characteristics per an EU AI Act-aligned prompt; 3 ranks candidates against a job description via a fine-tuned sentence-embedding model and cosine similarity; and 4 computes a nine-metric fairness suite spanning counterfactual score delta, mean absolute rank change, flip rate , group-fairness top-K retention, four-fifths/impact ratio , and merit-aware Recall@K, nDCG@K, equal opportunity, equalized odds families, each with bootstrap confidence intervals, significance tests, and Benjamini-Hochberg correction, culminating in an automated PASS/INVESTIGATE/FAIL report with a composite risk score. On an example corpus of 5 job orders, 100 base candidates, and 10 demographic treatments 90 metric x variant evaluations : score shifts, top-K retention, and merit-aware rate gaps stay within tolerance for every treatment, but a rank-stability metric MARC and nDCG@K each surface borderline findings - including one on the neutral baseline itself - that a score- or retention-only view would miss. The results argue for multi-metric, multi-family auditing over any single aggregate score, and for LLM-agent-generated audits as a practical, low-cost complement to human-curated audits for any candidate-job matching pipeline.