TL;DR
Get your next haul delivered free — and shop member deals
- Fast, free delivery on millions of items
- Access to Prime Big Deal Days deals on October 6–7
- Prime Video, Amazon Music and more included
Federal Reserve Governor Christopher Waller said AI could improve cross-border payment screening and routing, while AI agents may eventually shop and pay on consumers’ behalf. He described agentic commerce as being in an early phase, with autonomous purchases requiring stronger safeguards. The speech did not set a deployment timetable or endorse a particular payment system.
Federal Reserve Governor Christopher J. Waller said artificial intelligence could help improve cross-border payment screening and routing, while AI agents may eventually make purchases and payments for consumers. In a speech at Sibos 2026 in Miami on September 29, he described agentic commerce as early-stage and said more extensive trust mechanisms and guardrails would be needed before agents could shop autonomously at scale.
Waller said financial firms have already used machine learning to detect payment fraud and large language models for tasks such as payment reconciliation. He pointed to sanctions screening and anti-money-laundering systems as areas where language models could help interpret context and reduce false-positive alerts. That could leave investigators more time for complex, higher-risk cases and reduce delays for legitimate payments, he said. Waller framed these tools as an addition to faster, traditional anomaly-detection methods, not a replacement.
He also discussed AI’s possible role in cross-border payment routing, where providers weigh cost, speed and reliability while making currency-conversion and liquidity decisions. AI agents could help search among those options when given suitable data and criteria, Waller said. On cybersecurity, he acknowledged concern that AI may support more sophisticated attacks, while also identifying defensive uses such as threat detection, exposure management and finding software vulnerabilities.
Waller described two models of agentic commerce. In an agent-assisted purchase, an AI helps a customer find products but the customer makes the decisions and pays. In an agent-delegated purchase, the customer authorizes an agent to shop and pay within set limits. One example in the speech was an agent using a pre-funded virtual card to buy groceries based on a person’s past purchases. Waller said consumer-to-business transactions are expected to be an early use case, with assisted shopping preceding more autonomous delegation.
The shift from AI helping people find goods to AI carrying out purchases changes who initiates a payment. A customer may still set the budget, merchant limits or other conditions, but an agent acting independently creates questions about authorization, mistaken purchases and accountability. Payment providers and merchants would need reliable ways to identify the agent, verify its authority and apply the customer’s limits.
Waller’s remarks also connect AI adoption to the operation of existing payment networks. Better screening and routing could affect how quickly cross-border transactions are reviewed and completed. At the same time, the interconnected nature of payment systems makes cybersecurity failures consequential beyond a single company. Waller said defenders must account for a broad attack surface, while an attacker may need to exploit only one vulnerability.
AI-powered cross-border payment security tools
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
From Fraud Detection to Agentic Commerce
Waller’s speech followed a Federal Reserve industry roundtable on agentic commerce that he said took place at the beginning of summer 2026. He reported that participants broadly considered the market to be in an early phase, while recognizing that wider adoption could alter commerce and payments. His address placed that emerging use case alongside payment firms’ existing work on fraud detection and reconciliation.
The speech was delivered at Sibos 2026, a financial-services industry conference, and focused on questions about how AI might change payment systems. Waller discussed potential applications and risks; he did not announce a Federal Reserve program, new payment rule or approved product. The source report is a speech published by the Bank for International Settlements and identifies Waller as a member of the Federal Reserve Board of Governors.
virtual prepaid cards for autonomous shopping
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Safeguards and Adoption Remain Open
Waller outlined possible benefits and risks but did not provide evidence that AI systems are already delivering the described improvements across payment networks. The speech cited research showing that language models can reduce false-positive alerts, but it did not specify a result, test conditions or a common measure of performance. The size of any gains in screening, routing or currency conversion therefore remains unclear from the speech.
It is also unclear how quickly consumers will adopt delegated purchasing, which firms will provide the agents, or what technical and legal standards will govern authorization, disputed purchases and losses. Waller did not set a timetable for the trust mechanisms he said would be needed, or identify a single approach that payment firms should use.
AI fraud detection payment systems
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Industry Work on Trust and Controls
Waller’s speech points to further work by payment firms and other market participants on the conditions for agent-led purchases, including how customers grant authority and set spending limits. The speech did not announce a formal deadline or regulatory next step. Any expansion from assisted product search to autonomous shopping is likely to depend on whether companies can build controls that customers, merchants and payment providers can rely on.
For cross-border payments, providers will also need to test AI tools against existing screening and routing systems and evaluate both performance and security. Waller called for a deliberate approach, but did not name a specific deployment schedule. Further details about standards, adoption and measurable outcomes remain to be established.
As an affiliate, we earn on qualifying purchases.
Key Questions
What did Christopher Waller say about AI and payments?
Waller said AI could help with cross-border screening, cybersecurity and payment routing, and that AI agents may eventually shop and pay for consumers. He described agentic commerce as an early-stage development.
What is the difference between agent-assisted and agent-delegated commerce?
In the assisted model, an agent helps with product search while the customer chooses and pays. In the delegated model, the customer authorizes the agent to make purchases and payments within specified limits.
Did Waller announce a new Federal Reserve policy?
No. The speech discussed possible applications, risks and safeguards. It did not announce a new Federal Reserve rule, program or payment product.
What safeguards would autonomous shopping need?
Waller said delegated commerce requires more extensive trust mechanisms and guardrails because it carries a greater risk of unintended purchases. He did not specify a single technical standard or timetable.
Can AI already improve cross-border payments?
Waller described potential uses in screening and routing, and referred to research on reducing false-positive alerts. His speech did not establish the scale of real-world gains across payment networks.
Source: primary
Fall Picks
fall essentials
As an affiliate, we earn on qualifying purchases.
