Visa is opening its vast payment network to artificial intelligence agents that can shop and buy products on behalf of consumers, marking a significant expansion of AI into the $4.6 trillion global payments industry.
The financial services giant announced Thursday the release of new developer tools that allow AI agents to connect directly to Visa's payment infrastructure, enabling what the company calls "agentic commerce" — a system where AI bots handle everything from product discovery to checkout completion based on consumer preferences and spending limits.
The move puts Visa at the forefront of a technological shift that could fundamentally alter how people shop online. Rather than browsing websites and manually completing purchases, consumers would set parameters for AI agents that then autonomously find, evaluate, and buy products across multiple merchants.
"Soon people will have AI agents browse, select, purchase and manage on their behalf," said Rubail Birwadker, Visa's Global Head of Growth, in an exclusive interview with VentureBeat. "These agents will need to be trusted with payments, not only by users, but by banks and sellers as well."
The announcement escalates an intensifying competition with rival Mastercard, which launched its own AI agent payment system called Agent Pay in April. Both companies are racing to capture what industry analysts expect will be a rapidly growing segment of digital commerce as AI agents become more sophisticated and widely adopted.
New developer tools slash integration time from weeks to hours
Visa's new offering centers on two key products: a Model Context Protocol (MCP) Server that provides secure access to Visa's payment APIs, and the Visa Acceptance Agent Toolkit, which allows both technical and non-technical users to deploy AI-powered payment workflows using plain language commands.
The MCP Server represents a significant technical breakthrough, providing AI agents with a standardized way to communicate with Visa's trusted network without requiring custom integrations for each application. Developers can now move "from idea to functional prototype in hours instead of days or weeks," according to the company.
For Visa, which processes over 300 billion transactions annually across 200 countries, the expansion into AI commerce is both defensive and offensive. The company faces growing pressure from fintech startups, real-time payment networks, and tech giants looking to disrupt traditional payment rails.
"The way commerce has worked over the past 30 years is you are physically in a merchant environment — you go to a Sephora or Best Buy — or you're in an online merchant environment," Birwadker explained. "But in an agentic environment, consumers are going to remain in these AI platforms, and they're going to want to buy goods, pay for goods or services."
How Visa plans to prevent AI bots from going rogue with your money
The shift to AI-powered commerce introduces unprecedented security and trust challenges. When consumers delegate purchasing decisions to software, traditional fraud prevention methods require significant adaptation.
Visa has implemented multiple layers of protection, including immediate tokenization of card credentials, device-specific authentication, and what Birwadker calls "payment signals" and "payment instructions" that verify AI agent actions align with original consumer intent.
"Your PII or your PAN is never going to be exposed," Birwadker said, referring to personally identifiable information and primary account numbers. "We almost immediately take that pan, we convert it into a token, and we authenticate that token and tie it to a specific device for a specific application."
The company has also developed a matching process that prevents transaction completion until it confirms an AI agent's intended purchase matches what the consumer originally requested. This addresses concerns about AI "hallucinations" — instances where language models generate incorrect or nonsensical outputs.
However, Birwadker acknowledged that not all risks can be anticipated. "Look, we don't claim to have solved everything, this is a new form of commerce. A lot of things could happen that we haven't foreseen," he said. "We're letting the robot slowly and not letting a lot of water flow through the pipes until we actually learn and understand how this actually works."
Mastercard rivalry heats up as both networks chase AI commerce gold rush
The timing of Visa's announcement reflects broader urgency across the payments industry to position for an AI-driven future. Consumer adoption of AI platforms has accelerated dramatically, with platforms like ChatGPT, Perplexity, and others attracting hundreds of millions of users who increasingly use these tools for product research and recommendations.
According to a March 2025 Salesforce study, 66% of shoppers expressed interest in using AI agents to secure high-demand items before they sell out, while 65% want agents that can automatically purchase products when they reach target prices.
Mastercard's Agent Pay system launched just one day before Visa announced its Intelligent Commerce initiative in April, highlighting the competitive intensity. Both companies are working with major AI firms including OpenAI, Microsoft, Anthropic, and Stripe to develop the ecosystem.
"This isn't just another product," Birwadker said. "We've partnered with leading AI platforms to bring this to life, and we're helping the entire ecosystem build agents that are transacting safely."
The competitive pressure extends beyond traditional payment networks. PayPal has partnered with AI search engine Perplexity to enable automated purchasing, while Amazon is testing a "Buy for Me" feature that would allow AI agents to make purchases on behalf of users.
Visa's simple bet: more AI transactions mean more revenue
Despite the technological complexity, Visa's business model for AI commerce remains straightforward: the company profits when more transactions flow through its network using Visa-branded cards.
"As long as a Visa card is used, we are good," Birwadker said. "Our goal here consistently remains that consumers, developers can innovate and iterate on top of our APIs, because ultimately, when they do that, Visa cards become what consumers end up using."
This approach reflects Visa's position as a network operator rather than a direct merchant. The company facilitates transactions between parties rather than taking inventory risk or direct customer service responsibility.
Early testing has been promising, with Birwadker noting that Visa has "30 plus partners live in the sandbox" conducting friends and family testing. The company is also working with merchants and acquirers to ensure the backend infrastructure can handle AI-generated transactions at scale.
From invoice generation to revenue analysis: AI agents tackle business tasks
The Visa Acceptance Agent Toolkit demonstrates immediate practical applications for the technology. Businesses can use simple natural language commands to automate routine payment tasks.
"A very simple example would be, you know, can you go look at all my invoices I've paid, all the money that I've received, and create a revenue summary for me and figure out where the gaps are, and create invoices for the unpaid," Birwadker explained.
The toolkit can also automatically generate invoices from Excel files containing payment information, using natural language processing to convert data into formatted billing documents ready for transmission.
Such capabilities could significantly reduce administrative overhead for small and medium-sized businesses, potentially accelerating adoption among Visa's merchant partners.
Consumer trust and regulatory hurdles loom over AI payment future
The expansion of AI into payments raises important regulatory questions that remain largely unresolved. Different jurisdictions have varying requirements for payment authentication, data localization, and consumer protection that must be incorporated into AI agent systems.
"Every market has their local regulatory nuances in our payments work, and they can incorporate all of that," Birwadker said. "So from an agentic perspective, the payments work regardless of where a consumer is, regardless where the merchant is."
Consumer trust represents perhaps the biggest long-term challenge. While AI adoption has surged, surveys show mixed sentiment about delegating financial decisions to automated systems. A December 2024 eMarketer study found only 24% of U.S. consumers comfortable sharing data with AI shopping assistants.
Visa's decades-long track record in fraud prevention — the company says it prevents $40 billion in fraud annually — provides credibility, but AI commerce introduces new variables that traditional security models weren't designed to address.
The stakes for Visa couldn't be higher. After building a dominant position in traditional digital payments over three decades, the company now faces the challenge of maintaining that leadership as commerce shifts toward AI-powered automation. Success could solidify Visa's role as the backbone of digital transactions for another generation. Failure could open the door for nimbler competitors to reshape the payments landscape entirely.
As consumers increasingly delegate their shopping decisions to artificial intelligence, one thing seems certain: the companies that can make those transactions feel both effortless and secure will write the next chapter of digital commerce. For now, Visa is betting its network effects and fraud prevention expertise will prove just as valuable in an AI-driven world as they were in the era of plastic cards and point-of-sale terminals.