Fuelled by growing AI-enabled attacks, fraudulent transactions in digital banking and money transfer is expected to reach 2.2 billion by 2031.
This is the findings of a recent report by Juniper Research, highlighting how generative AI enables new and more adaptive scam typologies, while emerging agentic AI allows fraudsters to coordinate multi-stage attacks and adjust tactics dynamically as victims or controls respond.
The rise of generative and agentic AI is turning fraud into a more persistent and adaptive financial risk for enterprises. As fraudsters use AI to personalise scams, coordinate multi-stage attacks and circumvent controls, CFOs face greater exposure not only to direct financial losses but also to operational disruption, customer remediation costs, regulatory penalties and reputational damage.
This puts fraud prevention increasingly within the CFO’s broader responsibility for financial risk management and internal controls, as rather than treating fraud as a transactional banking issue, finance leaders need greater visibility into how fraud risks evolve across the customer and payment journey, and whether existing controls can adapt quickly enough.
AI can play a role in this by creating a continuous view of customer and transaction risk, bringing together behavioural, identity and payment signals.
For CFOs, the key question is increasingly whether investments in fraud and risk technology can reduce financial exposure, strengthen controls and demonstrate measurable returns, rather than simply detecting suspicious transactions after the fact.
For banks, it is imperative now to move beyond detecting suspicious transactions towards further understanding how risks are developing. Juniper Research recommends lenders to use AI to build a continuous view of customer risk; analysing how behavioural, identity and payment signals evolve together to enable intervention before fraudulent transactions are authorised.
