Cost of AI fraud has fallen by a hundredfold: Unico, Liminal report

A new study from Liminal and Unico offers yet another look at the new era of identity fraud, in which isolated incidents have evolved into a “coordinated, economically rational network amplified by generative artificial intelligence.”

According to the Identity Fraud Intelligence Report: Trends and Insights, sophisticated fraud assisted by generative AI already accounts for 23 percent of fraud attacks globally, leading to  losses exceeding $400 billion per year. Intermediate-level forgeries using manipulated image or video streams – including basic injection attacks – account for 45.8 percent of cases.

Hybrid methods combining deepfakes, injection methods and sophisticated physical manipulation represent 23.3 percent of the total. Deepfakes cause average losses exceeding $280,000 annually, and nearly 20 percent of cases result in losses above $500,000.

Conversely, the cost of mounting attacks has fallen more than a hundredfold, as generative AI tools become ever more freely available online via subscription models. Per the report, “the barrier to entry for sophisticated attacks has nearly disappeared.”

“In this environment, nearly 90 percent of fraud professionals identify synthetic content, including manipulated documents and bot impersonation, as one of their leading concerns today.” Identity has become “the central point at which revenue is either protected or lost.” And no market is safe.

Defense must meet fraud with coherent strategy

Davi Reis, vice president of strategy at Unico, says the findings “reveal a structural imbalance in the current fraud prevention system.”

“While criminals operate through coordinated networks, at low cost and at scale, institutions still rely on static and isolated defenses, increasing their exposure to repeated and sophisticated attacks,” Reis says. He calls the current moment a paradigm shift. “If fraud operates as a network, intelligence must also be interconnected – making biometrics essential to identifying repeat fraudsters across the digital ecosystem.”

Article Topics

AI fraud  |  digital identity  |  fraud prevention  |  generative AI  |  Liminal  |  synthetic identity fraud  |  Unico