Major payments and technology companies are rushing to build out the infrastructure they see as the next evolution in global commerce: artificially intelligent agents that can perform searches, compare prices, and make purchases on behalf of consumers. This trend, known as “proxy commerce,” reflects consumers’ increased reliance on chatbots for everyday tasks, such as searching for products and transactions online. However, until recently, these tools lacked important functionality. Shoppers could search and compare within the chatbot, but needed to leave the interface to complete their purchase. Payment giants like Visa and Mastercard say things are changing. Over the past year, the companies have been racing to build the systems and partnerships needed for the next steps in commerce, and initial pilots of the technology are already underway. Payment executives told CNBC that the technology could become a reality by 2026 and be more transformative than the rise of e-commerce platforms like Amazon. “A major shift in commerce occurred when payments moved from a primarily brick-and-mortar world to an e-commerce world,” said Sandeep Malhotra, executive vice president of Core Payments, Asia Pacific, Mastercard. “Now we are witnessing the next change, and that is the transition from the world of e-commerce to the world of agency commerce,” he said. “We went from cash to digital, and now we’re going from digital to intelligent.” Many of the details of where and how agent commerce will work are still being worked out, but the term generally refers to AI systems that discover products, compare deals, and complete payments within chatbots on behalf of users. This makes shopping more seamless by curating options based on your specific requests, instead of navigating through multiple websites and apps with traditional e-commerce. It will be interesting to see how companies adapt when the prevalence of price discovery and shopping becomes the norm rather than the exception. TR Ramachandran Payment executive, Visa’s head of APAC products and solutions, said one obvious early use case could be booking flights and vacations. For example, a user might ask an AI commerce agent, “Find me the cheapest red-eye flight nonstop from Singapore to Tokyo for under $500.” Agents will be able to scan, offer options, book tickets, and pay using the user’s saved payment credentials, all within the chat interface. Mastercard’s Malhotra said the technology also allows shoppers to authorize agents to make purchases even when they are offline, such as automatically purchasing an item if its price falls below a pre-set threshold. Early pilots Visa and Mastercard have deployed initial frameworks to secure bot-driven transactions and have already completed pilot programs with selected users and merchants. TR Ramachandran, Visa’s head of products and solutions for Asia Pacific, told CNBC that commercial use of personalized and secure agent transactions could be realized as early as the first quarter of 2026. Ramachandran said the ground for agent commerce is fertile, as more than half of Visa’s overall transaction volume is already done through e-commerce, and data shows demand for AI to assist with shopping. A December Visa study found that nearly half of U.S. shoppers use AI to improve their shopping experience when finding gifts and comparing prices. Meanwhile, AI-driven traffic to U.S. retail sites grew 4,700% year over year in July, according to Adobe research. Agent commerce transactions are expected to occur through AI platforms commonly used by consumers, such as ChatGPT and Google’s Gemini, as well as merchant, bank, and app-specific agents. Companies like Mastercard and Visa are working closely with AI giants like OpenAI to prepare for the transition. In September, OpenAI launched a “Buy with ChatGPT” feature that enables instant checkout within the platform. Meanwhile, Perplexity partnered with PayPal to roll out free shopping products for US users in November. Concerned about potential price pressures and the loss of direct access to customers, major retailers are also testing agent commerce tools on their own. Amazon began testing Buy For Me earlier this year in an effort to block external AI agents from crawling its website. Potential Issues Despite early momentum, agent commerce still raises a number of potential structural, security, and liability concerns. A major focus for payments companies is the creation of so-called “agent tokens,” which use cryptographic authentication to verify authorized AI agents from human users and distinguish them from malicious bots. For example, Visa launched its “Trusted Agent Protocol” in October with Cloudflare to create cryptographically authenticated records for transactions initiated by bots. Ramachandran said Visa also plans to add “payment signals” for banks, provide more transaction details and strengthen agent authentication with behavioral intelligence. Another obvious issue is liability if the AI agent makes a mistake. For example, if you buy the wrong color bike or book a hotel room for the wrong night. Traditionally, disputes involve four parties: the consumer, the issuing bank, the acquiring bank, and the seller. “There is now a fifth player in the value chain. AI platforms are being integrated into the value chain because customers want to be there,” Ramachandran said. “You almost have to expect mistakes to happen and put guardrails and protections around them,” he added. These potential issues require stronger guardrails and permissions, as well as a robust dispute system, and these challenges are still being resolved in the pilot phase. Big Impact Proponents say agent commerce saves time, reduces search costs and gives consumers more access to information and transactions. “I can only see the benefits of agent commerce. Consumers will have better access to information, better access to products, better access to services and better experiences,” Malhotra added. But as AI-driven price discovery and changes in consumer behavior become more commonplace, merchants may face pressure to adapt. Visa’s Ramachandran said: “It will be interesting to see how companies adapt when price discovery and shopping ubiquity become the norm rather than the exception.” Payment professionals expect merchants to implement agent authentication, create their own AI agents to interact with consumer agents, offer loyalty programs, and redesign upsell strategies. Despite challenges and unknowns, payment companies say the transition to agent commerce is inevitable. “There’s not a lot of clarity on exactly when it will expand,” Ramachandran said. “But based on our experience and adoption of the overall (large-scale language model) platform, it could take months, not years.”
