Proxy commerce is taking shape within payment network platforms, with earnings calls and materials showing progress in moving from concept to controlled implementation.
The focus during these quarterly updates has been on how AI agents work within existing payment credentials, authorization flows, and risk frameworks.
Visa CEO Ryan McInerney Transformation framework operationally, explained throughout the company Profits Assume that “AI and agent commerce will expand our addressable market,” linking this expansion to the growth of transactions and new use cases across consumer and commerce flows.
American Express and MasterCard Deliver similar signals. American Express is included Agent capabilities Within its closed network, while MasterCard is expanding its open network through… Coding and credential formatting.
Extend existing rails
The networks build proxy commerce as an extension of the payment infrastructure rather than a parallel system.
Visa described a model in which agents can research, decide and ultimately pay on behalf of consumers and businesses, with transactions executed using direct cards and existing merchant acceptance points. This approach allows Visa to scale agent-driven payments without requiring new acceptance infrastructure.
Advertisement: Scroll to continue
“Our strategy as a Visa-as-a-service group will help us drive future growth in important ways,” McInerney said during the call, including proxy commerce alongside stablecoins and money movement.
Mastercard seeks to achieve the same result with its own structure. CEO Michael Maybach “The ecosystem continues to evolve,” he said during the company’s earnings call, adding that Mastercard is “engaged in shaping what comes next with key players, including Google, Microsoft, OpenAI And other partners.”
American Express, operating under a closed-loop model, has introduced an Agentic Commerce Experiences development kit designed to integrate its cards directly into AI-driven transactions, signaling a more vertically integrated approach.
Trust and control
The biggest challenge, as noted during the calls, is ensuring that these transactions can be trusted, authenticated, and reversed when necessary.
To address this, Visa is relying on tokenization, real-time data, and network-level visibility, as demonstrated on Wednesday (April 29), when the company announced the expansion of its Agent ready Program for Latin America and Asia. McInerney stressed the advantage of scale on Visa’s quarterly call, saying Visa processes “more than 300 billion transactions annually,” providing the data needed to manage risk in automated environments.
Mastercard offers additional controls at the credential level. The company launched “verifiable intent,” described as a tamper-resistant record of user authorization for agent-initiated payments, with the goal of resolving disputes and clarifying liability.
American Express addresses the same problem with security frameworks. Management described it to her Agent purchase protection As an “industry first commitment” to support transactions conducted by registered agents, indicating that liability rules are being defined along with technology.
High-frequency and complex payment flows
The networks do not expect an immediate shift in the consumer payment process. Instead, they point to specific areas where automation can increase transaction volume.
Visa highlighted B2B payments as a primary use case, noting that agents can automate payment initiation directly from invoices and contracts. The embrace of agent commerce also emphasizes the ability to manage approvals independently. This reduces friction in workflows that are already digital but still require manual intervention.
Mastercard has linked agent commerce to its broader expansion into value-added services. CFO Sachin MehraForeign Policy’s commentary showed that value-added services and solutions revenues rose 18% on a currency-neutral basis, driven by demand for security, authentication, and analytics capabilities that support automated transactions.
Miebach emphasized the strategic direction, saying that “diverse services powered by our data and how we approach partnerships” are key to enabling new payment experiences.
American Express cited operational efficiency gains, with management noting that AI initiatives are already delivering measurable productivity benefits, positioning agent commerce as a long-term driver of growth and customer engagement.
In all three networks, the pattern is consistent. Agent trade is expected to increase transaction intensity, particularly in cross-border and trade flows, while reshaping everyday consumer behaviour.
For all of our PYMNTS AI coverage, subscribe to our daily newsletter Amnesty International newsletter.
