AI-enabled fraud to hit 2.2bn banking transactions by 2031

Fraudulent transactions in digital banking and money transfer are on course to exceed 2.2 billion annually by 2031, up from 773.7 million in 2025. That is the headline finding from a new report by Hampshire-based tech research firm Juniper Research, which frames the acceleration not as an incremental worsening of existing fraud vectors but as a structural shift in how financial crime is orchestrated and funded.

The driving force, Juniper argues, is the collapsing cost of mounting personalised attacks. Generative AI now allows fraudsters to produce high-quality, targeted social engineering campaigns at industrial scale. The result is that end-user customers, rather than banking infrastructure itself, have become the primary target. Compromising a payment system requires technical sophistication; convincing an authenticated customer to authorise a transfer does not.

Agentic AI raises the automation ceiling

The report's most consequential near-term warning concerns agentic AI, the class of systems capable of pursuing multi-step goals autonomously and adapting their behaviour in response to feedback. Juniper identifies this as a qualitative leap beyond generative fraud tools: where generative AI lowers the cost of content production, agentic systems can coordinate entire attack sequences, pivoting tactics in real time as victims or detection controls respond.

The practical implication is a compression of the window available for banks to intervene. As report author Shane O'Sullivan observes: "Coupled with a reduced cost to commit fraud, AI enables fraudsters to create and adapt attacks faster, while instant payments reduces the window that banks have to identify suspicious behaviour. Banks must therefore combine behavioural, identity and payment intelligence prior to authorisation, rather than relying on controls after a payment has been initiated."

The shift from post-payment detection to pre-authorisation intelligence is not merely a technical upgrade; it is a reorientation of the entire fraud-prevention operating model, with significant implications for compliance architecture, data infrastructure investment, and vendor procurement.

Convergence pressure on financial infrastructure and AI spend

The Juniper findings sit at the intersection of two macro trends that Disrupts readers will recognise. First, the rapid productisation of foundation models has made sophisticated AI capabilities accessible to actors well outside the traditional technology supply chain, lowering barriers to criminal innovation in the same way it lowers barriers to commercial innovation. Second, the global rollout of instant payment rails, from the UK's Faster Payments and the EU's SEPA Instant to India's UPI and Brazil's Pix, has removed the settlement buffer that historically gave fraud teams time to flag anomalies.

Together these forces are reshaping capital allocation inside major banks. Investment in legacy, rules-based fraud detection systems is increasingly difficult to justify when the attack surface is evolving faster than static rule sets can be updated. The growth trajectory implicit in Juniper's figures, if borne out, will drive material procurement shifts toward real-time behavioural analytics platforms, identity intelligence vendors, and AI-native monitoring layers. Fraud prevention, long treated as a compliance cost centre, is being reframed as a competitive differentiator in retail and corporate banking alike.

For cross-sector investors, this also reframes the fraud-prevention vendor landscape as a direct beneficiary of the same AI diffusion that is threatening the incumbents those vendors serve. The market for detection and prevention tooling is structurally tied to the growth of the threat it counters, a dynamic that tends to support durable revenue expansion regardless of broader credit cycle conditions. Juniper's report covers 44,700 datapoints across a five-year forecast period and includes a competitor leaderboard, providing institutional buyers and capital allocators with a more granular view of vendor positioning as the market consolidates.