Banks have plenty of fraud technology, but getting all of it to work as a cohesive line of defense is becoming a problem.
The PYMNTS Intelligence report “Scams Get Personal: Why Banks Pay More to Fight Fraud” found in September that 78% of financial institutions said the complexity of new technology systems was their top barrier to innovation, up from 48% a year earlier. It was the largest increase among the 14 barriers tracked in the report.
The finding came after years of adding tools designed to identify suspicious customers, devices and transactions. Behavioral analytics, device fingerprinting and enhanced knowledge-based authentication reached 100% adoption among surveyed institutions in 2026. Machine learning and artificial intelligence were used by 96%, as were rules-based algorithms.
Adding another detection capability, however, isn’t the same as improving the overall defense. The data showed that institutions were stacking overlapping systems, making the connections among those systems a growing operational issue.
Fraud Moves Across the Stack
Integration becomes more important as fraud becomes harder to isolate at a single checkpoint.
Scams in which the account holder was manipulated into initiating a payment accounted for 22.4% of fraud dollars in 2026, up from 12% in 2025. Relationship or trust fraud more than doubled from 6% to 12.6%, while account takeover through compromised credentials rose to 28.2% of fraud dollars.
The report found that 50.6% of fraud dollars involved either a legitimate login or manipulation of the customer, up from 27.6% two years ago. Physical forgery and counterfeit fraud fell 66% to 6.1% of fraud dollar value, and impersonation of the account holder declined 37% to 10.8%.
We’d love to be your preferred source for news.
Please add us to your preferred sources list so our news, data and interviews show up in your feed. Thanks!
The changes put pressure on fraud systems to connect signals that can originate at different points in the customer and payment journey. A valid credential doesn’t necessarily mean the person using it hasn’t been compromised, and a payment initiated by the actual customer isn’t necessarily legitimate if a scammer manipulated the customer into sending it.
Financial institutions appear to be responding by seeking greater control over how their fraud technology works together.
According to the report, 71% of financial institutions said they plan to develop new fraud detection systems in-house, up from 44% in 2025. Plans to outsource fraud detection and prevention fell to 29% from 51%.
Yet “in-house” rarely means proprietary from end to end.
Among institutions developing systems internally, 93% combined third-party fraud technology with their own proprietary tools. Just 0.2% described their approach as entirely their own, and none of the institutions surveyed reported independently building any of the 11 fraud technologies tracked in the report.
The result is a tech-based challenge that moves beyond simply deciding whether to buy or build. Institutions still rely on outside technology while trying to control how the products interact with internal data, rules and systems.
Machine learning showed some evidence of progress. Among institutions using it, 86% said it enabled them to operate proactively by catching problems before they happen, up from 25% in 2025. Large banks reported that benefit at 93%, followed by neobanks and FinTechs at 91%, and regional and community banks at 83%.
At PYMNTS Intelligence, we work with businesses to uncover insights that fuel intelligent, data-driven discussions on changing customer expectations, a more connected economy and the strategic shifts necessary to achieve outcomes. With rigorous research methodologies and unwavering commitment to objective quality, we offer trusted data to grow your business. As our partner, you’ll have access to our diverse team of PhDs, researchers, data analysts, number crunchers, subject matter veterans and editorial experts.
Add as Preferred Source
