Key takeaway: Juniper Research forecasts that AI-powered social engineering and agentic AI coordination will push global fraudulent digital-banking and money-transfer transactions up 180% by 2031. The shift moves fraudsters’ focus away from bank infrastructure and toward end users and their authenticated payment sessions.
Juniper Research forecasts that fraudulent transactions in digital banking and money transfer will exceed 2.2 billion globally by 2031. That figure is up 180% from 773.7 million in 2025, according to Digital Transactions.
Juniper attributes the growth to AI, which lowers the cost of running large-scale, personalized social engineering attacks. The firm says this trend makes bank customers what it calls increasingly viable fraud targets.
The firm also points to agentic AI. It says agentic AI lets criminals coordinate multi-stage attacks and adjust them in real time as victims and bank controls respond. Juniper says this makes fraud more persistent, more personalized, and harder to identify.
Juniper’s forecast includes a regional breakdown. It projects 132.6 million fraudulent transactions in North America by 2031 — the sixth-largest total among the eight regions the report covers, according to Digital Transactions.
Juniper’s report, as cited separately by other outlets, also projects 175.4 million fraudulent transactions across Africa and the Middle East by 2031. Report author Shane O’Sullivan said a reduced cost to commit fraud, combined with AI, lets fraudsters create and adapt attacks faster. He added that instant payments shrink the window banks have to catch suspicious behavior before funds move.
Juniper recommends that financial institutions build a continuous view of customer risk. Institutions should track how behavioral, identity, and payment signals change together so they can flag a fraudulent transaction before authorization.
Why it matters: The forecast points to a structural shift in where banking fraud happens. Fraud is moving away from attacks on core banking infrastructure and toward end users during their own authenticated sessions.
For merchants, processors, and fraud teams, that shift suggests a growing share of digital-banking and payments losses will come from account takeover and social-engineering-driven transfers. That marks a change from losses driven mainly by card-present or card-not-present fraud.
Source: Digital Transactions