Hong Kong – November 09, 2026 -- Deutsche Bank and Ant Digital Technologies have jointly published a white paper, Agentic Payments in Wholesale Finance, that advocates "bounded agentic execution" rather than unrestricted autonomy for AI agents handling institutional payments.
White paper names three wholesale use cases for AI agents
The paper identifies treasury and liquidity management, cross-border payment orchestration and trade finance as the areas with the most operational and strategic value.
In treasury, agents could forecast liquidity shortfalls and surpluses and execute compliant internal transfers, investments, external financing and foreign exchange operations. In cross-border payments, they could weigh FX rates, payment rails and execution timing to optimise cost, speed, regulatory compliance and settlement risk. In trade finance, they could interpret trade documentation, flag data discrepancies, coordinate exception handling and generate financing or payment instructions once conditions and approvals are met.
Three-layer IMF architecture separates AI reasoning from transaction execution
The paper draws on a reference architecture proposed by the International Monetary Fund (IMF):
- Intent and Orchestration: adaptive agent reasoning, decision-making and workflow orchestration.
- Control and Authorisation: deterministic, rule-based constraints that keep actions within predefined authorisation boundaries.
- Settlement: final execution of financial transactions.
Safeguards within the framework include explicit, bounded and revocable authorisation limits, robust agent identity binding and Know Your Agent (KYA) verification protocols.
Ant's Anvita Flow platform serves as the implementation case study
The paper features Anvita Flow, an agentic platform developed by Ant Digital Technologies, as an example of orchestrating complex payment workflows in real-world operating environments while keeping tight control over agent authority and execution.
Zhuoqun Bian, President of Ant Digital Technologies, wrote in her foreword that decision-making, control and execution must be separated so that an agent "may propose but cannot extend its own authority." Sabih Behzad, Head of Digital Assets and Currencies Transformation at Deutsche Bank, contributed the other foreword.
The full white paper is available for download from the websites of both institutions.