Agentic payments in wholesale finance

Artificial intelligence (AI) is beginning to move beyond generating information and recommendations. Increasingly, AI agents can interpret objectives, assess alternatives and take action. For wholesale finance, this raises a fundamental question: How much authority should institutions delegate to AI agents?

A new white paper from Deutsche Bank and Ant Digital Technologies explores the emerging concept of agentic payments and what it could mean for wholesale finance. As Sabih Behzad, Head of Digital Assets and Currencies Transformation at Deutsche Bank, puts it in his foreword: “The potential value lies not simply in executing payments automatically, but in improving how those decisions are made, coordinated, and controlled.”

The paper examines three priority use cases:

• Treasury liquidity management, where agents could anticipate liquidity shortfalls and surpluses and assess permitted responses across internal transfers, investment, external funding, and foreign exchange.
• Cross-border payment orchestration, where agents could evaluate interdependent decisions across FX conversion, payment rails, and execution timing, based on cost, speed, availability, regulatory requirements, and settlement risk.
• Trade finance, where agents could interpret transaction data and documents, identify discrepancies, orchestrate controls and exceptions, and prepare financing or payment instructions once the required conditions and approvals are satisfied.

So where could agentic payments deliver real value? And what safeguards are needed before they can move from experimentation towards institutional adoption?

Download the white paper to explore how agentic payments could reshape decision-making across wholesale finance.