Mastercard and Google Team Up to Build Trust for AI-Powered Shopping

Verifiable Intent creates a tamper-resistant, cryptographic record of what a user authorized when an AI agent acts on their behalf.

Mastercard has unveiled Verifiable Intent, a new open, standards-based trust framework co-developed with Google, designed specifically for “agentic commerce” — a world where artificial intelligence (AI) systems dont just assist shoppers, but actively plan, decide, and complete purchases autonomously.

The core problem Verifiable Intent aims to solve is visibility: when a consumer delegates a purchase to an AI agent, the clear “click buy” or “tap to pay” moment that traditionally signals intent disappears. Mastercards Chief Digital Officer Pablo Fourez argues that this creates a new challenge for every party involved — consumers need assurance their instructions were followed, merchants need confirmation an agent is authorized to buy, and issuers need to distinguish legitimate activity from fraud.

To address this, Verifiable Intent creates a tamper-resistant, cryptographic record of what a user authorized when an AI agent acts on their behalf — linking identity, intent, and action into a single, privacy-preserving audit trail.

The framework uses Selective Disclosure, a privacy control technique, to ensure that only the minimum necessary information is shared between parties and only when needed, allowing merchants and issuers to verify transactions without access to sensitive consumer data.

It leverages widely adopted standards from the FIDO Alliance, EMVCo, the Internet Engineering Task Force, and the World Wide Web Consortium, and is designed to work across agentic protocols, devices, wallets, and platforms. Mastercard says Verifiable Intent will be integrated into its Agent Pay APIs in the coming months.

Crypto Rails Join the Fray

Not everyone sees traditional payment networks as the right foundation for AI-driven commerce, however, highlighting a growing debate about whether AI agents will ultimately transact through incumbent networks like Mastercard or bypass them entirely in favor of crypto-native infrastructure.

“Very soon there are going to be more AI agents than humans making transactions. They cant open a bank account, but they can own a crypto wallet. Think about it,” Coinbase CEO Brian Armstrong posted on X today.

In September, EigenCloud, Ethereums largest restaking protocol with nearly $9 billion in total value locked, announced a partnership with Google Cloud to serve as the verifiable backbone for AI agent payments.

Meanwhile, the Ethereum Foundation launched a dedicated AI initiative called the dAI Team, with a stated mission to make Ethereum the preferred settlement and coordination layer for the emerging “machine economy.”

The following month, attention turned to x402 protocols, which enable AI agent payment systems and increase the practicality of agentic AI-led finance.

Taken together, these developments paint a picture of an industry racing to solve the same core problem from two very different directions. Mastercard and traditional finance are building trust layers on top of existing payment rails, while crypto proponents are betting that blockchain infrastructure is better suited to a world where AI agents are first-class economic actors.