AI and banking & payments: the intelligence layer

In banking and payments, AI disproportionately reinforces execution, risk, and control layers, while accelerating commoditization of front-end and orchestration-only models. This is supported by structural protections specific to the sector: regulatory and licensing architecture, card scheme certification and banking charters, decades of proprietary transaction and case data, and deep embedding in mandatory compliance and operational workflows. Fraud, compliance, onboarding and office-of-the-CFO are example areas where AI adoption is scaling.

Agentic commerce is projected to reach $2tn to $5tn by 2030, representing 3% to 7% of global payment volumes.

J.P. Morgan, 2026 Payments Outlook

Our view is that AI value in banking and payments accrues to platforms that combine regulatory standing, proprietary data, and embedded workflows. The full perspective sets out what makes the sector structurally distinct, the pace of agentic adoption today, where value is shifting across the value chain, and the risks to that view.

Download to read the full perspective, including case studies on how our portfolio companies Splitit, Backbase, finperks and Cordant are embedding AI.