Last updated: September 2026. Techeconomix Editorial Team — researched using primary reporting from Deloitte’s Center for Financial Services, Javelin Strategy & Research, Mastercard, and Thomson Reuters Institute. See “Sources & Methodology” at the end of this article.
Quick answer: Global financial crime losses crossed $579.4 billion in 2025, according to the Nasdaq Verafin Global Financial Crime Report, and generative AI is now the primary force reshaping both sides of the fight — powering both more convincing scams and the AI-driven defenses banks use to catch them. Deloitte projects gen-AI-enabled fraud losses in the US could hit $40 billion by 2027, up from $12.3 billion in 2023. Here’s what’s actually driving the numbers, and the specific scam patterns showing up most often right now.
The Scale of the Problem
The headline figures are sobering. The FTC recorded $15.9 billion in US consumer fraud losses in 2025, a record high, with investment scams accounting for $5.7 billion of that total, according to Hedge Think’s analysis of the FTC data. Globally, identity fraud losses exceeded $50 billion in 2025, and early 2026 indicators suggest that figure will be surpassed before year-end. Mastercard’s 2025 research found organizations lost an average of $60 million each to payment fraud over the prior year.
Photo by Sora Shimazaki via Pexels
Why AI Changed the Fraud Landscape
Generative AI has industrialized deception, according to Michal Tresner, CEO of ThreatMark, speaking to Thomson Reuters Institute. Fraudsters now use AI to run automated attacks at scale, generate synthetic identities that bypass traditional verification checks, and craft social-engineering scripts convincing enough that victims willingly authenticate and move their own money — even while every individual security control shows green. That last point matters: much of today’s fraud isn’t a technical breach at all. It’s a manipulated customer acting on fraudulent instructions.
A specific and rapidly growing pattern illustrates this well: bank-impersonation text messages are now the most-reported text-message scam, up nearly 20-fold since 2019, with a typical victim losing around $3,000, according to consumer-finance research cited by myfinancialfreedomtracker.com. The scam works by sending a text that looks exactly like a real bank fraud alert; when the recipient replies, a “fraud agent” calls within seconds with caller ID even showing the bank’s real name, walking the victim through moving money into what’s described as a “safe account” — which is, of course, the scammer’s account.
New Account Fraud Is Surging
Javelin Strategy & Research’s 2026 annual identity fraud study, titled “The Illusion of Progress,” found a 31% rise in new account fraud, affecting 5.4 million victims and causing $7 billion in losses. That’s notable because it moved in the opposite direction of other fraud categories: romance scams declined 41% in the same period, and overall identity fraud losses stabilized. The study’s authors describe this as evidence that AI is reshaping which fraud types grow and which shrink, rather than uniformly increasing all fraud everywhere.
How Banks Are Fighting Back
The same AI capabilities powering fraud are now central to detecting it. Modern fraud-detection systems function as what industry analysts call “agentic defense networks” — continuously analyzing transactions, flagging anomalies in real time, and escalating suspicious activity autonomously before losses occur, according to Emburse’s 2026 banking AI guide. American Express reported improving fraud detection by 6% using advanced machine-learning models that analyze sequences of transactions rather than single events in isolation.
Mastercard’s research found 83% of industry leaders say AI has reduced false positives and customer churn — a meaningful shift, since older rule-based fraud systems were notorious for blocking legitimate transactions and frustrating customers as often as they caught real fraud. The stakes for getting detection speed right are especially high with instant payments: unlike traditional card transactions that can sometimes be reversed, transfers over networks like FedNow and RTP, which we’ve covered in our piece on FedNow’s 2026 adoption numbers, settle in seconds and are difficult to claw back once sent.
What This Means for You
- Be suspicious of urgency: Real banks rarely pressure you to move money to a “safe account” over the phone within minutes of a text alert.
- Caller ID can be spoofed: A call showing your bank’s real name is not proof the caller is legitimate.
- Verify independently: If you get a suspicious fraud alert, call your bank back using the number on your card or official app, not a number provided in the text or by the caller.
- New accounts need extra scrutiny: Given the 31% surge in new-account fraud, be cautious with any financial product opened in your name that you don’t recognize, and monitor your credit reports regularly.
Frequently Asked Questions
How much is AI-driven fraud costing US consumers?
The FTC recorded $15.9 billion in US consumer fraud losses in 2025, a record high. Deloitte projects gen-AI-enabled fraud losses could reach $40 billion in the US by 2027.
What is the most common AI-powered bank scam right now?
Bank-impersonation text messages are currently the most-reported text scam, up nearly 20-fold since 2019, typically followed by a spoofed “fraud agent” phone call.
Is new account fraud increasing?
Yes. Javelin Strategy & Research found a 31% rise in new account fraud in its 2026 study, affecting 5.4 million victims and $7 billion in losses.
How are banks using AI to fight fraud?
Banks increasingly use AI systems that continuously monitor transactions in real time and flag anomalies before money moves, rather than relying on static, rule-based checks after the fact.
Sources & Methodology
This article draws on data and reporting from: the Nasdaq Verafin Global Financial Crime Report (2025); Deloitte’s Center for Financial Services research on generative AI and banking fraud; Javelin Strategy & Research’s 2026 identity fraud study “The Illusion of Progress”; Mastercard’s 2025-2026 payment fraud prevention reports; Thomson Reuters Institute’s 2026 AI-powered fraud trends analysis; and Emburse’s 2026 AI fraud detection guide. Figures reflect the most recently published data as of this article’s last-updated date.
This article is for informational purposes and does not constitute financial or security advice. If you believe you’ve been targeted by fraud, contact your financial institution directly using verified contact information.
