68% of banks turn to AI as fraud outpaces rules-based systems

This is a shift at the heart of the State of Fraud and Financial Crime in the U.S. 2025 report produced by PYMNTS Intelligence in partnership with Block. The report, based on a survey of 200 executives at U.S. financial institutions, found that fraud has not only become more sophisticated but also more focused on credential theft and account takeovers.

At the same time, agencies are investing more aggressively in artificial intelligence (AI), behavioral analytics and cloud infrastructure to keep pace.

Fraud losses at the industry level remain contained, but pressures are uneven and growing, particularly for large banks and digital-first businesses.

Key data points from the report illustrate the changing landscape:

• Currently, 71% of fraud incidents and financial losses originate from schemes run by unauthorized parties, a significant increase from 48% last year.

• The average fraud loss rate rose to 0.8 basis points, with large banks reporting losses of more than 3.5 basis points, more than four times the survey average.

• 68% of financial institutions have increased spending on fraud detection year over year, while 46% report that fraud schemes are increasing in sophistication.

The resurgence of unauthorized access reflects a broader evolution in tactics. Fraudsters are exploiting credentials, manipulating payment information and targeting faster payment channels. As the report's breakdown of fraud types shows, digital payments fraud and credential compromise account for a growing share of transaction volume and dollar value. This is not a static crime. It will adapt. Institutional response. The cycle repeats.

However, the impact extends beyond the balance sheet.

Data shows that half of financial institutions say fraud hurts customer loyalty. Forty-four percent believe this has caused damage to the brand and reputation. Nearly as many reported lost business opportunities. Today, fraud is both a growth issue and a risk issue.

Technology is moving from optional to required. Machine learning and behavioral analytics are becoming foundational tools. 70% of organizations say AI enables them to blend active and passive defenses.

About eight in 10 fintech companies and large banks report implementing advanced behavioral analytics. Still, about one in five institutions, especially smaller and regional banks, operate without these capabilities. This gap is widening.

Agencies also face overlapping pressures. Regulatory compliance challenges were cited by 47%. 41% believe payment options are faster or more diverse. Nearly half said the sophistication of fraud itself is increasing. The result is a complex operating environment where compliance, modernization and fraud prevention must occur simultaneously.

Looking ahead, investment plans point to continued momentum. Half of institutions plan to expand their use of cloud-based fraud platforms. 51% of respondents want to increase outsourcing. Forty percent are investing in deep learning systems. Fraud prevention is becoming multi-layered and collaborative.

The report's conclusions are clear. Fraud will continue to evolve as fast as the systems designed to stop it. Institutions that combine predictive analytics, real-time data, and human oversight will be best positioned to maintain trust in an increasingly digital economy.

Fraud is no longer just a cost of doing business. This is a test of resilience.

At PYMNTS Intelligence, we work with businesses to uncover insights and drive intelligent, data-driven discussions about changing customer expectations, a more connected economy, and the strategic shifts needed to deliver results. With rigorous research methods and an unwavering commitment to objective quality, we deliver data you can trust to grow your business. As our partner, you'll have access to our diverse team of PhDs, researchers, data analysts, digital numerators, subject matter veterans, and editorial experts.