Criminal attacks involving artificial intelligence and specialized malware are putting banks and their customers under mounting pressure. As recent surveys by security researchers show, AI-driven deception tactics such as deepfakes have already caused billions of dollars in losses worldwide, while mobile trojans simultaneously target account credentials on smartphones.
Billions in Losses from Cloned Voices and Synthetic Videos
According to a study published in July 2026 by cybersecurity firm Surfshark, global losses from deepfake fraud have reached at least $3.7 billion. Nearly 90 percent of this total occurred between early 2025 and mid-2026. Deepfakes are audio and video recordings manipulated or synthetically generated using generative AI. Cybercriminals use this fake material to bypass internal company approval workflows or customer identification checks by impersonating executives or account holders.
The number of unreported cases is believed to be substantial: According to the study, fewer than five percent of voice-cloning fraud victims report the incident to authorities at all. A KPMG survey indicates that 62 percent of bank managers now rank deepfakes among their greatest security risks. Researchers from Deloitte also warn that annual losses from AI-driven financial fraud in the US could surge to up to $40 billion by 2027.
Android Trojans Hijack One-Time Passwords
Alongside AI-based deception, smartphones are becoming an increasingly prominent target. Reports from spring 2026 show a 56 percent surge in Android banking trojans for the year 2025. The number of malicious installation packages even jumped by 271 percent. Security researchers are focusing particularly on the Mamont and Creduz malware families.
Attackers primarily distribute their malware to mobile devices disguised as everyday utility or reward apps. Once installed, the programs demand extensive permissions. In doing so, the malware can intercept one-time passwords (OTPs) for transactions and autonomously hijack active online banking sessions.
High Costs for the Financial Sector
Institutions’ infrastructure is also under siege. According to an IBM study released in mid-2026, a data breach cost the financial industry an average of $5.56 million in 2025—roughly 25 percent above the cross-industry average. While the overall time to identify incidents decreased thanks to defensive AI systems, attackers now deploy AI tools themselves in around 16 percent of all attacks.
How Consumers Can Protect Themselves
As criminals increasingly rely on mobile devices and identity spoofing, banking customers should follow key security practices:
- Review permissions: Do not grant newly installed apps access to SMS, notifications, or accessibility services unless strictly necessary for the app’s intended functionality.
- Exercise caution with reward and utility apps: Download mobile applications only from trusted sources and avoid suspicious installation packages.
- Be skeptical of phone-based verifications: Because attackers can convincingly clone voices, suspicious transaction requests should always be cross-checked via an official, separate communication channel with the bank.
Sources: Borncity.com
