AI Unleashes New Breed of Fraudsters, Socure Report Warns

Socure, a prominent identity-verification firm serving 18 of the top 20 US banks, is set to release its inaugural Fraud Insights report on October 19. This comprehensive study leverages data from Socure’s extensive consortium network, encompassing over 20 billion transactions, 620 million accounts, and 178 million unique identities, which represents approximately 72% of the adult US banked population. Supplementing this robust dataset is a global survey of 1,253 consumers conducted via the Prolific research platform.

The report’s headline conclusion reveals a structural shift in fraud dynamics: artificial intelligence (AI) is not merely enhancing the productivity of existing fraudsters but is actively recruiting individuals who would not have engaged in fraudulent activities otherwise. This new demographic is labeled the “AI-Fluent Fraudster.”

A striking characteristic of the AI-Fluent Fraudster is the absence of prior fraudulent behavior in over half of this cohort when AI tools are not involved. Their baseline fraud rate, at 4.1%, is statistically similar to the general population’s 5.0%. This poses a significant challenge for fraud and identity teams, as traditional behavioral signals, typically built around repeat offenders and established fraud patterns, are unlikely to flag this group before their initial fraudulent act.

The problem is compounded by the scale at which AI-Fluent Fraudsters operate. They commit approximately three times as many fraudulent acts as traditional offenders, leading to an estimated six-fold increase in total fraud volume when combining breadth and frequency. A notable 50.6% of AI-Fluent Fraudsters committed at least one fraudulent act ten or more times, a repeat-offense rate 5.5 times higher than that of traditional fraudsters.

Their motivational profile is also distinct. Curiosity serves as the primary driver for 41.8% of AI-Fluent Fraudsters, who seek to test AI's capabilities rather than acting out of financial desperation – a rate nearly double that of traditional fraudsters. The report draws a parallel to Napster, where downloading music felt psychologically detached from shoplifting a CD. Similarly, 15.8% of AI-Fluent Fraudsters reported feeling no personal responsibility when using AI to generate fake hardship letters or falsified bank statements, a figure nearly three times higher than among traditional fraudsters. Demographically, the typical AI-Fluent Fraudster is male and under 30, with Gen Z representing the fastest-growing segment within this cohort.

These findings arrive at a critical juncture as fraud liability frameworks are tightening across both sides of the Atlantic. In the UK, reimbursement rules for authorized push payment fraud shifted material liability toward payment firms in late 2024, escalating the commercial stakes for accurate detection. The EU’s revised Payment Services Directive (PSD3) extends similar logic to member states. Concurrently, the US Consumer Financial Protection Bureau faces ongoing pressure to clarify liability standards for first-party fraud, a category directly addressed by this report.

For compliance and fraud-operations teams, Socure's data sharpens a difficult question: if this new cohort of AI-enabled fraudsters is behaviorally indistinguishable from legitimate customers during onboarding, then existing verification and monitoring controls, designed around known fraud rings and synthetic identity networks, are at best insufficient. Identity verification providers, fraud-tooling vendors, and the financial institutions relying on them will need to urgently revisit and update their model training sets to accurately reflect the AI-Fluent profile at scale.