American Express Launches Intent Intelligence for AI Purchases
American Express introduces the Intent Intelligence framework, a cryptographic system that verifies AI-driven shopping intents, helping brands prevent…
Carlos Martínez 4 min read
Executive summary
- The shift: American Express officially launched its “Intent Intelligence” framework on October 13 2026, creating a technical standard for verifying purchases made by AI agents [Business Wire].
- The mechanics: The system uses a “Create Intent” API to generate an Intent ID and a cryptographically encoded Proof of Intent token, locking down specific merchant, budget and product constraints before any transaction executes [Tech Times].
- The adoption curve: 66 % of U.S. consumers say they will likely use AI agents to find the best price within 12 months, according to Amex Trendex data [Value Add VC].
- The strategic angle: This is not just a fraud‑prevention tool; it is a new layer of trust infrastructure that brands must integrate to remain visible in agentic commerce flows.
- The action item: If you are relying on legacy checkout flows, you are building for a dead end. The ACE Developer Kit is already available for integration [Business Wire].
Table of contents
The days of assuming a human is clicking the “Buy Now” button are ending fast. For decades, retail logic was built on the idea of a conscious decision‑maker staring at a screen, weighing options and committing. That model is fracturing. When an AI agent executes a purchase, the psychological friction disappears, but the verification logic collapses with it. Who authorized this? Did the agent stick to the parameters? American Express just answered that question with a technical release that signals a hard turn in the payments industry.
On Tuesday, October 13 2026, Amex published its Intent Intelligence framework. It was presented at TechCrunch Disrupt 2026 in San Francisco by Hannah Bozian, Vice President of Agentic Partnerships and Strategy. The timing is not accidental. It coincides with a massive shift in consumer behavior where trust is no longer implicit in the card number; it must be explicit in the code.
The anatomy of a verified intent
The core innovation here is not about blocking bad bots. It is about encoding the why behind the transaction. The framework utilizes an API called “Create Intent.” When a shopper engages with an AI agent, this API generates two critical artifacts: an Intent ID and a Proof of Intent token [Tech Times].
Think of the Proof of Intent token as a cryptographic receipt for the shopper’s desires. It encodes specific restrictions: the exact merchant, the maximum budget, and the required product characteristics. If the AI agent tries to upsell to a more expensive model or redirects the shopper elsewhere, the token invalidates the transaction. It is a hard stop.
This changes the dynamic between brands and payment processors. Previously, if a customer charged back a purchase made by an agent, the burden of proof was murky. Now, the intent is verifiable data. Luke Gebb, Head of Global Innovation at Amex, framed this as a way to capture and authenticate the customer’s purchase intention before the transaction executes [Business Wire]. For a CTO, this means your backend needs to speak this language. If you cannot parse the Intent ID or validate the token, you are invisible to a significant portion of high‑value agentic traffic.
Why “verification” is actually a retention strategy
There is a common misconception that this framework is solely about fraud reduction. While that is part of the value, the bigger story is about customer control. When a shopper uses an agent, they are delegating trust. If that delegation feels risky, they won’t use the agent. If they feel their budget and preferences are cryptographically locked in, they will use it.
Consider the contrast with the current state of agentic commerce. We have seen cases where AI agents abandon carts due to friction or lack of clarity. As discussed in our analysis of why AI agents abandon retail shopping carts, the lack of clear handshakes between the agent and the merchant creates drop‑off. The Intent Intelligence framework solves this by making the handshake explicit. It removes the ambiguity that causes agents to bail.
This is particularly relevant for brands that have seen success with conversational assistants. For example, an AI assistant used by a major home‑goods retailer has managed to triple shopper purchases by streamlining the interaction. However, those interactions still relied on traditional checkout flows that lack explicit intent verification.
FREE SESSION
See Epinium’s AI services → – free 30‑min diagnostic
SERVICES BY EPINIUM
Book free diagnostic → – free 30‑min diagnostic