American Express has introduced a new technical framework designed to help card issuers verify purchases made by artificial intelligence-powered shopping agents, marking a significant step forward in the rapidly evolving landscape of digital commerce.

The payments giant released its Intent Intelligence framework, outlining a standardized method for recording a shopper’s precise purchase instructions before automated AI systems execute transactions on their behalf. As autonomous shopping assistants gain traction, the system aims to provide card issuers with the verifiable data necessary to detect potential fraud, streamline payment approvals, and resolve transaction disputes when unexpected issues arise. For retailers, the framework offers a reliable method to determine whether an ordering mistake originated from an AI agent or a system error on the merchant’s end.

Although American Express has confirmed that Intent Intelligence does not yet feature live integrations, the company’s decision to publish the framework is intended to encourage broader adoption across the wider payment and technology ecosystem rather than immediately introduce new consumer-facing capabilities.

“Providing authorized purchase intent before an agent transacts can give issuers greater confidence that an agent is acting as a customer intended, while also helping issuers proactively manage risk and reduce the chance of costly disputes across the payments ecosystem,” said Luke Gebb, head of global innovation at American Express, in the company’s announcement.

The rollout arrives as financial institutions and technology firms aggressively develop infrastructure for agentic commerce. In this emerging paradigm, advanced AI agents are capable of autonomously searching for products, comparing prices across multiple platforms, and finalizing purchases without requiring direct human intervention at every step of the checkout process.

Despite the technological momentum, consumers remain notably cautious about ceding financial control to autonomous software. A consumer survey of 1,000 U.S. adults released in August by payments company NMI revealed that 58% of respondents harbor concerns that an AI shopping agent might mistakenly purchase the wrong item, size, color, or brand. Furthermore, 60% of surveyed consumers expressed anxiety that an agent could easily exceed their intended spending limits.

Although American Express initially introduced Intent Intelligence as part of a broader suite of services in April, the recent publication makes the complete technical architecture accessible for other industry participants to build upon. The company stated that its ultimate goal is to help advance a cohesive, universal industry approach for managing consumer intent within the agentic commerce space.

What the New American Express Framework Means for Retailers

For traditional retailers and e-commerce platforms, AI-driven purchasing errors introduce unprecedented logistical and customer service challenges. A merchant could theoretically fulfill an incoming order precisely as transmitted by an automated system, only to face a contentious chargeback or dispute because the AI agent purchased an item the human shopper never actually requested or authorized.

This is precisely where documented records of original shopper instructions, broadly defined as purchase intent, serve as a critical safeguard. By capturing what the customer explicitly approved prior to execution, the framework provides transparent evidentiary trails showing whether the AI agent genuinely adhered to those pre-set parameters.

Intent Intelligence constitutes one of five core services initially introduced by American Express in April through its Agentic Commerce Experiences (ACE) developer kit. Company representatives noted that the framework differs significantly from existing implementations because it transmits a customer’s authenticated purchase intent to the card issuer much earlier in the transaction lifecycle—specifically before the AI agent executes the final payment.

American Express indicated that Intent Intelligence supports a versatile array of customizable purchase controls. Shoppers retain the ability to establish strict maximum spending thresholds, definitive expiration dates, and specific merchant preferences. Furthermore, users can designate exact timing conditions for when an agent is permitted to make a purchase, such as waiting until an out-of-stock product returns to inventory or executing a transaction precisely during Cyber Monday promotions.

The architecture is also engineered to accommodate complex, multi-layered purchasing scenarios. For instance, a consumer could authorize an agent to buy supplemental flight insurance only if the system successfully secures the corresponding airline tickets first. Additionally, it handles interconnected multi-transaction orders, such as purchasing matching shoes and socks from entirely separate online retailers.

Designed for operational flexibility, Intent Intelligence works seamlessly across diverse agentic commerce models regardless of which entity initially captures the consumer’s instructions. Depending on the specific architectural configuration of an agent, those instructions can be transmitted directly to a card issuer, routed through a digital wallet provider, or managed via another certified payment intermediary.

Why AI Shopping Errors Create New Payment Risks

Beyond assisting issuers with transaction verification, the framework directly supports broader strategic initiatives by American Express to safeguard both shoppers and merchants against liabilities stemming from autonomous AI errors.

When American Express initially unveiled its ACE developer kit, it outlined prospective coverage referred to as Agent Purchase Protection. This proposed coverage would apply specifically to eligible transactions involving registered AI agents utilizing the developer kit and properly transmitting authenticated purchase instructions. For merchants, the long-term objective is to systematically reduce chargebacks, friction, and disputes.

However, American Express confirmed that these protections are not yet commercially active, noting that there are currently no live ACE integrations utilizing Intent Intelligence, which remains a mandatory prerequisite for qualification. In the interim, cardmembers continue to rely on the standard protections and benefits inherently associated with their eligible American Express payment cards.

This emerging complexity also influences how consumers attribute responsibility when automated shopping transactions go wrong. In the August NMI survey, 34% of consumers indicated they would hold the retailer primarily accountable if an AI tool completed a purchase prematurely. Another 24% stated they would blame their payment provider, while an overwhelming 62% pointed directly at the AI software provider as the ultimate responsible party.

Where American Express Fits in Agentic Commerce

American Express is far from alone in racing to establish robust security and infrastructure standards for AI-driven commerce. Major financial networks and technology enterprises have launched competing protocols to secure and streamline automated transactions.

In April 2025, both Visa and Mastercard introduced dedicated frameworks enabling developers to securely bridge AI agents with their legacy payment networks. Visa’s Intelligent Commerce program provides developers with specialized application programming interfaces and tools to integrate core payment functions directly into conversational AI environments, supported by global partnerships with major technology players including OpenAI, Microsoft, Anthropic, Stripe, and Samsung.

Simultaneously, Mastercard introduced Agent Pay, a technology designed to empower AI agents to securely execute transactions on behalf of users across both consumer retail and business-to-business environments.

Prominent technology companies are simultaneously advancing their own proprietary and open standards. OpenAI and Stripe previously co-developed the Agentic Commerce Protocol to allow AI systems to securely interface with merchants for order initiation. Meanwhile, Google introduced the Agent Payments Protocol, which utilizes digitally signed cryptographic records to document shopper authorizations, drawing early adoption from platforms like Salesforce, Shopify, Cloudflare, and Etsy.

Google expanded its footprint further with the Universal Commerce Protocol, an open standard developed alongside Shopify and other industry partners to facilitate seamless checkout experiences directly within AI shopping interfaces. Major retailers, including Newegg, Wayfair, Walmart, and Macy’s, have progressively adopted these integration standards to allow consumers to complete transactions natively within generative AI environments like Google’s AI Mode and Gemini.