AI Powered Resume Fraud Detection

Abstract

Resume fraud misrepresentation or fabrication of qualifications, work experience, or skills on CVs and resumes poses a growing threat to organizations, leading to poor hiring decisions, financial loss, and reputational damage. This paper presents an automated Resume Fraud Detection System that combines natural language processing (NLP), information extraction, and machine learning to detect in consistencies and likely fabrications in candidate resumes. The System extracts structure dentities (degrees, institutions, dates, companies, job titles, skills), cross-validates them against External authoritative sources and intra resume consistency rules, and uses supervised and anomaly-detection models to score the likelihood of fraud. Evaluation on a curated dataset containing verified and synthetic fraudulent resumes demonstrates the effectiveness of the proposed approach, achieving high precision in flagging suspicious resumes while maintaining acceptable recall. The proposed system aims to reduce manual screening time, standardize fraud detection, and integrate with HR pipelines for scalable pre-hire verification.