In Brief: The rate of AI-enabled fraud attempts has increased by over 180 percent; more than half of all documented financial fraud cases now involve active AI deployment.
Cybercriminals are using generative AI to scale fraud attacks—from forged invoices to deceptively authentic deepfake video conferences. Organizations are responding with real-time anomaly detection in their payment systems.
The Sumsub Identity Fraud Report documents a sophistication shift in the threat landscape: cybercriminals combine stolen identity data with artificially generated biometric features and forged documents. The spectrum ranges from automatically generated supplier invoices to customized social engineering campaigns that all formally fit into traditional validation processes.
Multiple incidents demonstrate the operational risk of this technology. In Hong Kong, attackers transferred around 25 million US dollars via deepfake video conference by synthetically reconstructing the image of the Chief Financial Officer and other executives in real time. In March 2025, a finance director in Singapore authorized payments of 499,000 US dollars during an alleged Zoom call with forged participants. Technology companies such as LastPass and Wiz were attacked using cloned CEO voices based on YouTube conference recordings, which distributed forged instructions via WhatsApp.
Traditional rule-based controls fail in the face of this dynamic, as criminal transactions pass all static thresholds and validation steps flawlessly. As a countermeasure, security leaders are integrating forensic AI systems directly into payment processing workflows. These platforms perform millisecond-precision real-time auditing and use behavioral signals and statistical anomaly detection to identify suspicious patterns.
According to the Deloitte Banking and Capital Markets Outlook, damages caused by AI in the financial sector have already reached double-digit billions and are projected to reach 40 billion US dollars by 2027. This underscores the urgency of examining payment flows based on behavioral patterns rather than only on formal plausibility.
Source: www.it-daily.net · Published 17 July 2026
Lumi AI News — AI-assisted curation in accordance with Article 50 EU AI Act. Paraphrase and classification by Lumi News Pipeline v1.7.3.