Artificial intelligence is evolving beyond mere chatbots, with tech companies promising personal assistants capable of performing tangible tasks according to users’ preferences. However, as of now, these AI agents haven’t significantly transformed daily routines. Visa is aiming to change this by integrating your credit card information into these AI systems. This integration will allow AI agents—advanced versions of systems like ChatGPT—to purchase items such as clothing, groceries, or travel tickets on your behalf based on your budget and preferences.
“This could be crucial,” states Jack Forestell, Visa’s chief product and strategy officer. He likens its potential impact to the revolutionary advent of e-commerce. On Wednesday, Visa revealed collaborations with AI chatbot creators, including companies like Anthropic, Microsoft, OpenAI, Perplexity, and French startup Mistral. This partnership aims to connect their AI systems to Visa’s payment network and is supported further by industry giants such as IBM, digital transaction pioneer Stripe, and global phone maker Samsung. Initial projects start this week with broader implementation anticipated next year.
Visa envisions these AI capabilities evolving from a futuristic concept to a practical tool for managing mundane shopping tasks. Over the past months, they’ve been working on overcoming technical challenges so that everyday users can adopt this technology. For AI firms, Visa’s support could provide a competitive edge against tech giants like Amazon and Google that currently dominate the digital commerce space and are developing their own AI-assisted platforms.
The tech sector often showcases what it refers to as agentic AI, yet its presence in everyday life is limited. Most of these agents are merely adaptations of large language models, the foundation for engaging chatbots that can draft emails, summarize content, or assist with programming. These systems can analyze extensive datasets and offer product recommendations but struggle with facilitating the purchase process.
“The initial phases of agent-based commerce are effective at the browsing and suggestion stage but falter when it comes to executing payments,” notes Forestell. He adds that Visa aims to bridge this gap by enabling AI systems to access funds needed for transactions, a complexity AI platforms have been unable to tackle independently.
Visa’s new AI-focused initiative was unveiled almost a year after significant changes were introduced to US credit and debit card operations, signaling a move towards making physical cards increasingly obsolete. As consumers grow accustomed to handling payments digitally through tools like Apple Pay, Visa envisions AI agents undergoing a verification process to act on behalf of users, ensuring transaction legitimacy and resolving any disputes.
According to Forestell, AI agents won’t entirely hijack the shopping experience but might prove valuable for repetitive tasks like purchasing groceries, home hardware, or even planning holiday gift lists. Complex errands such as booking flights could benefit greatly from a digital assistant’s autonomy. However, for more experiential purchases such as luxury items, AI agents may play a supportive, more subtle role.
As for concerns regarding credit card debt—given U.S. consumers’ credit card balances peaked at $1.21 trillion last year, according to the New York Federal Reserve—Forestell assures that spending boundaries set by users will maintain human oversight over purchase decisions. Initially, AI agents will confer with buyers for approval of transactions like booking a flight ticket, but could ultimately be authorized to “spend a specific amount on any airline to fulfill travel requirements,” he suggests.
One attractive aspect for AI developers in teaming up with Visa is access to a user’s transaction history, granted through explicit user consent. This capability can enhance recommendation accuracy, such as suggesting the top laptops based on past purchasing behaviors, as noted by Dmitry Shevelenko of Perplexity, a company currently allowing its chatbot to book hotels and make select purchases. Though in nascent stages, Shevelenko sees vast potential in AI commerce, much of which is captured by emerging collaborations like the one sought by Perplexity and others, including OpenAI, as they eye potential acquisitions such as Google’s Chrome browser amid ongoing antitrust discussions.