AI-driven identity fraud is becoming a growing concern in South Africa as criminals move away from traditional scams and toward more sophisticated attacks powered by generative technologies. Sumsub’s 2025–2026 Identity Fraud Report shows that although global identity fraud rates have declined slightly, AI-based impersonation and document forgery are rising sharply.
The report records a 180 percent year-on-year increase in what it categorizes as “sophisticated fraud,” which includes deepfakes, synthetic identities and autonomous agents designed to mimic human behavior. At the same time, overall global fraud rates fell from 2.6 percent in 2024 to 2.2 percent in 2025, suggesting stronger verification systems and tighter regulatory oversight are reducing lower-level scams.
“Fraudsters are moving from high-volume, low-effort scams to targeted, AI-powered operations. Meanwhile, businesses and regulators innovate with advanced verification and behavioral analytics,” said Hannes Bezuidenhout, Sumsub vice president for sales in Africa.
South African businesses are reporting similar patterns. Criminals increasingly exploit AI tools during high-activity periods such as the festive season, using phishing, social engineering and account takeovers to breach accounts. The report notes that companies must now adopt verification systems capable not only of authenticating people, but also of detecting when AI agents are acting on their behalf.
Sumsub warns that AI-driven fraud is likely to accelerate in 2026 as autonomous agents and synthetic identity networks become more advanced. Companies that adopt multi-layered verification systems and behavioral analytics will be better equipped to protect customers and limit financial losses.
