Deutsche Bank and Ant Group Release White Paper Exploring AI Agent Financial Payments

On November 9, Deutsche Bank and Ant Group jointly released a white paper titled "Intelligent Agent Payments in Wholesale Finance," exploring the application potential of AI Agents in areas such as institutional fund management, cross-border payments, and trade financing, along with governance and risk control frameworks. The white paper indicates that AI Agents can autonomously predict liquidity gaps, assess fund allocation and foreign exchange trading schemes, coordinate cross-border payment paths, handle trade financing documents, identify data discrepancies, and generate payment instructions. Related applications are expected to enhance the automation level and operational efficiency of complex business processes in financial institutions. In terms of risk control, the white paper advocates for a "limited authorization agent execution" model, proposing a reference architecture composed of three layers: intention and orchestration, control and authorization, and settlement execution. Core safeguarding measures include clear and revocable authority boundaries, verifiable agent identities, and a Know Your Agent verification mechanism. The white paper also lists the intelligent agent platform Anvita Flow developed by Ant Group as a practical application case.