The vision of artificial intelligence acting as a true personal assistant—one that doesn’t just answer questions but takes meaningful action in the world on our behalf—has been prominently featured in tech industry promises for months. While companies have enthusiastically marketed the concept of AI “agents” that understand our needs and can perform real tasks independently, the reality has fallen considerably short of these ambitious projections. These next-generation assistants have largely remained confined to information retrieval and text generation, stopping short of completing real-world transactions that would make them truly useful in our daily lives.
Visa, the global payments technology company, has identified this limitation as an opportunity for innovation and announced a groundbreaking initiative that could fundamentally transform how consumers interact with AI assistants. By providing these AI systems access to payment credentials, Visa aims to bridge the critical gap between recommendation and transaction, potentially revolutionizing digital commerce in the process.
Visa’s Vision for AI-Powered Commerce
On Wednesday, Visa publicly announced partnerships with several leading AI companies to integrate payment processing capabilities directly into their systems. This collaborative effort includes prominent U.S.-based AI developers Anthropic, Microsoft, OpenAI, and Perplexity, along with French AI company Mistral. The San Francisco-based payment processing giant is also working with IBM, online payment processor Stripe, and smartphone manufacturer Samsung on this initiative.
Jack Forestell, Visa’s chief product and strategy officer, emphasized the potential significance of this development in an interview, stating, “We think this could be really important. Transformational, on the order of magnitude of the advent of e-commerce itself.” This bold comparison underscores Visa’s belief that AI-facilitated commerce represents a fundamental shift in how consumers will shop in the future.
The initiative launches immediately with pilot projects, though Visa anticipates more widespread implementation will occur throughout 2025. This measured rollout reflects the complexity of integrating payment systems with emerging AI technologies while ensuring security and user trust remain paramount.
Bridging the Gap Between AI Capabilities and Real-World Action
The technology industry has produced numerous demonstrations of AI agents’ potential capabilities, showcasing systems that can search the web, make recommendations, and generate sophisticated content. However, few of these capabilities have translated into practical, real-world applications that extend beyond information retrieval and content generation.
Most current AI systems remain enhanced versions of large language models—the technology powering chatbots like ChatGPT that can draft emails, summarize documents, and assist with programming tasks. While these systems can effectively search the internet and recommend products, they typically hit a roadblock when it comes to completing actual transactions.
Forestell identified this limitation precisely, noting that “The early incarnations of agent-based commerce are starting to do a really good job on the shopping and discovery dimension of the problem, but they are having tremendous trouble on payments.” He described how AI assistants currently reach a point where they must defer back to the human user, essentially saying, “OK, you go buy it.”
This handoff creates friction in the user experience and undermines the promise of truly helpful AI agents. Visa believes its payment infrastructure can solve this problem by providing AI systems with secure, trusted access to financial resources needed to complete purchases—with appropriate guardrails and user authorization.
“The payments problem is not something the AI platforms can solve by themselves,” Forestell explained. “That’s why we started working with them.”
Building on Digital Payment Evolution
Visa’s new AI initiative represents a logical progression in the company’s broader strategy to adapt its payment infrastructure for the digital age. This announcement comes approximately one year after Visa revealed significant changes to how credit and debit cards function in the United States—changes designed to make physical cards and their 16-digit numbers increasingly obsolete as transactions move into digital environments.
Many consumers have already grown accustomed to digital payment systems such as Apple Pay that transform smartphones into payment instruments. Visa envisions a similar process for authorizing AI agents, with robust digital credential verification ensuring that transactions conducted by AI assistants on a customer’s behalf meet the same security standards as traditional payments. Importantly, Forestell emphasized that Visa’s existing frameworks for handling disputes will extend to AI-facilitated transactions, providing consumers with familiar protections.
Practical Applications and User Control
Rather than completely removing humans from the shopping experience, Forestell described a more nuanced vision for how AI agents might integrate into consumer purchasing habits. He suggested that AI-facilitated shopping would likely be most valuable for routine purchases that many people find tedious—such as groceries, home improvement supplies, and holiday shopping lists—or complex transactions like travel bookings where navigating numerous options can become overwhelming.
In these scenarios, consumers might prefer an agent that “just powers through it and automatically goes and does stuff for us,” Forestell said. The AI assistant could efficiently compare options, consider user preferences, and complete transactions without requiring constant human intervention.
However, Forestell acknowledged that many shopping experiences—particularly for luxury items—serve as a form of entertainment themselves, with customers enjoying the process of exploring options and making comparisons. In these situations, he envisions AI agents operating more in the background, providing assistance when needed but not taking over the entire experience.
Addressing Concerns About Consumer Debt
With American consumers already carrying substantial credit card debt—$1.21 trillion at the end of last year according to the Federal Reserve of New York—the prospect of AI agents making purchases raises legitimate questions about potential impacts on spending behavior and financial wellbeing.
Forestell addressed these concerns by emphasizing the importance of user control. He explained that consumers will establish clear spending limits and conditions for their AI agents, ensuring humans retain ultimate authority over their finances. Initially, AI agents will likely seek explicit approval before finalizing purchases, such as confirming a specific airline ticket. As the technology matures and user comfort increases, these agents might gradually receive more autonomy within carefully defined parameters—for example, authorization to “go spend up to $1,500 on any airline to get me from A to B.”
Data Integration and Personalization
Beyond streamlining transactions, Visa’s partnerships with AI developers offer additional value through potential data integration. With appropriate user consent, AI agents could access information about past credit card purchases to deliver more personalized recommendations.
Dmitry Shevelenko, Perplexity’s chief business officer, highlighted this benefit: “Visa has the ability for a user to consent to share streams of their transaction history with us. When we generate a recommendation—say you’re asking, ‘What are the best laptops?’—we would know what are other transactions you’ve made and the revealed preferences from that.”
Perplexity, a San Francisco-based AI startup, has already implemented some commerce capabilities in its chatbot, including hotel booking and other purchase types. However, Shevelenko acknowledged these features remain in the early stages of development. The company has also made headlines recently by indicating, along with OpenAI, that it would consider purchasing Google’s Chrome browser if a U.S. antitrust case forces the tech giant to divest parts of its business.
Strategic Implications for the Tech Ecosystem
For emerging AI companies, Visa’s support could provide a significant competitive advantage in the evolving landscape of digital commerce. By partnering with Visa to enable seamless transactions, these AI developers may improve their position against tech giants like Amazon and Google, which already dominate e-commerce and are developing their own AI agents.
This collaboration potentially creates a more diverse ecosystem of AI shopping assistants, giving consumers greater choice rather than concentrating commerce capabilities within a few dominant platforms. By providing essential payment infrastructure, Visa positions itself as a key enabler of this new generation of AI-powered commerce.
Looking Toward the Future
As these pilot projects begin and expand throughout 2025, both the technology and consumer behavior around AI shopping assistants will likely evolve considerably. Early implementations may focus on simple, low-risk transactions with substantial human oversight, gradually expanding to more complex purchasing decisions as the technology proves reliable and users become more comfortable delegating shopping tasks.
The ultimate success of this initiative will depend not only on the technical implementation but also on how effectively these AI agents balance convenience with maintaining appropriate user control and transparency. Consumers will need to develop new mental models for interacting with AI assistants that can spend money on their behalf—establishing boundaries, preferences, and comfort levels with this novel form of delegation.
While Visa’s announcement represents a significant step toward realizing the promise of truly helpful AI agents, the practical impact on everyday consumer behavior remains to be seen. If successful, this initiative could indeed mark a transformation in digital commerce comparable to the rise of e-commerce itself—fundamentally changing how people discover, evaluate, and purchase products and services in an increasingly AI-mediated world.
Acknowledgment: This article was written with the help of AI, which also assisted in research, drafting, editing, and formatting this current version.
