AI and Deepfake-Powered Fraud Skyrockets Amid Identity Fraud Stagnation

Sumsub’s Identity Fraud Report (Nov 25, 2025) finds that global identity fraud attempts slightly declined to 2.2% in 2025, but highly sophisticated attacks surged 180%, shifting fraud from high-volume to fewer, more damaging incidents that combine synthetic identities, AI/deepfakes, layered social engineering, and device/telemetry tampering. The report details first-party (synthetic IDs, chargeback abuse) and third-party (identity theft, account takeover) trends, highlights industries most affected (dating/media, financial services, crypto, professional services), and shows geographic variance with declines in Western countries and rises in the Middle East, APAC, Latin America and Africa; it recommends faster anomaly detection, behavioral analytics, and reducing manual prevention processes.