Neobanks, digital-only banks, and FinTech firms have significantly increased their spending on fraud prevention, which rose by 357% to $8 million per institution compared to $1.8 million previously. Despite these investments, average fraud losses grew by over 250%, climbing from $800,000 to $2.84 million.
Data from the PYMNTS Intelligence report indicates that the fraud-loss rate for these institutions decreased by 20%, falling from 0.90 to 0.72 basis points. However, the nature of threats is shifting, with over half of all fraud losses now involving customer manipulation or the use of legitimate login credentials.
Scams involving manipulated customers accounted for 22.4% of fraud losses in 2026, up from 12% the prior year. Account takeovers using compromised credentials now represent 28.2% of losses. These figures mark a significant increase from two years ago, when such methods accounted for 27.6% of total fraud dollars.
To combat these threats, all surveyed institutions have adopted behavioral analytics and device fingerprinting, while 96% utilize machine learning. Among those using machine learning, 91% report that it supports a proactive strategy, a substantial increase from 19% in 2025.
The impact of fraud extends to consumer behavior, as 28% of households that experienced a scam reported reducing or stopping their use of neobank accounts. This trend presents a dual challenge for digital providers, who must manage rising operational costs while maintaining customer trust and engagement.
