The Next Payments Revolution? Machines Shopping, Comparing And Paying On Their Own
By CU Today Staff —
NEW YORK--For nearly six decades, Mastercard has built its business around a simple challenge: creating trust between parties that have never met. In the age of artificial intelligence, the company believes that challenge is becoming even more important as software agents begin making purchases on behalf of consumers and businesses.
In an interview with PYMNTS CEO Karen Webster, Sherri Haymond, Mastercard's executive vice president and global head of digital commercialization, said the company's new Agent Pay platform is designed to provide the trust infrastructure needed for what many see as the next evolution of commerce.
Unlike traditional eCommerce, where consumers make every purchasing decision themselves, agentic commerce allows AI-powered software to search for products, compare options and execute transactions with minimal human involvement. Haymond told PYMNTS that consumers are increasingly embracing these models, with shopping activity already emerging across platforms such as Google and social media channels operated by Meta. As product discovery shifts from websites and apps into AI chat interfaces, the question becomes how consumers can safely delegate spending authority to software.
That trust challenge represents both the industry's greatest opportunity and its greatest risk. Haymond noted that when consumers hand purchasing authority to an agent, they are no longer personally reviewing product descriptions, delivery terms or return policies.
"The most important ingredient is trust," she told Webster. "If consumers, or people acting on behalf of businesses, don't trust that the system will work as intended, they're just not going to use it."
Mastercard's strategy is to build Agent Pay on top of its existing payments infrastructure, including tokenization, credential management and digital identity technologies that have been developed over the past two decades.
The company is also extending the concept beyond consumers through Agent Pay for Machines, a platform designed for machine-to-machine commerce. Haymond described a scenario in which a small business owner asks an AI assistant to establish an online presence. The software could purchase a domain name, build a website, launch a marketing campaign and compare service providers before presenting recommendations to the owner for approval. Once authorized, payments could be completed automatically using funding sources ranging from cards and virtual cards to bank accounts, lines of credit and stablecoins. In that environment, payments become embedded directly within business workflows rather than functioning as separate transactions.
Webster noted that such capabilities could dramatically lower barriers for small businesses that previously lacked the staff, time or resources needed to launch digital operations. Tasks that once required months of work and significant expense could potentially be completed in minutes, creating new opportunities for entrepreneurs and smaller organizations. Haymond characterized the development as both a solution to existing inefficiencies and the creation of an entirely new commercial category.
As with previous transformations in payments, Mastercard believes industrywide standards will determine how quickly agentic commerce scales. Haymond said standardization is essential because foundational infrastructure must operate consistently across platforms and providers. That philosophy drove Mastercard's development of Verifiable Intent, a framework designed to link identity, instructions and transaction outcomes into a verifiable record. Rather than keeping the framework proprietary, Mastercard turned it over to the FIDO Alliance earlier this year in an effort to encourage broader industry adoption.
Beyond facilitating payments, Mastercard sees the emergence of AI agents as creating entirely new revenue opportunities. As agents increasingly search for services, evaluate options and complete transactions independently, they become a new class of customers and buyers. Haymond said businesses may ultimately generate revenue not only from payments themselves but also from security services, transaction validation and data-driven personalization tools that help agents deliver better outcomes. In that model, agents evolve from simple assistants into a new commerce channel with its own economic ecosystem.
Looking ahead, Haymond told PYMNTS that the next 18 months will likely be defined by efforts to resolve competing technical standards and create smoother agent-driven workflows. She expects today's fragmented environment to give way to greater interoperability as industry participants work together to establish common rules. If that occurs, consumers and businesses may increasingly view AI-powered purchasing as a natural extension of commerce. For Mastercard, the objective remains the same one that has guided the company since its founding: creating the trust framework that allows strangers—or now, machines—to transact with confidence.
Originally reported by CU Today.