About this series: This is the opening piece of a four-part series examining the rise of agentic commerce and its legal implications. The series covers what agentic AI is and how it differs from earlier forms of artificial intelligence, who bears liability when autonomous systems transact on behalf of consumers, and how payment networks are responding through new authentication protocols. We begin where the technology begins: with the consumer.
Summer is approaching. Imagine opening your shopping app and typing: "Plan a 7-day trip in the Amalfi Coast for two adults and one kid; flights, a family-friendly hotel under $400 a night, close to the water."
Within seconds, an AI agent searches across platforms, compares flight routes and seat availability, reviews hotel cancellation policies, checks whether your loyalty points apply, selects the optimal payment method, and prepares both reservations. You receive a single notification asking for final approval.
You did not browse a single website. You did not read the fine print. The agent did. Your only act was a single click. What sounds futuristic is quickly becoming commercially viable. Major technology and payment companies are already building the infrastructure for what is being called "agentic commerce" — a model in which autonomous AI systems search, negotiate, and transact on behalf of users. Google unveiled a major AI-powered shopping ecosystem featuring an autonomous personal shopping agent. Mastercard launched Agent Pay, tokenized credentials that allow AI agents to transact securely on behalf of consumers, and is actively working with Microsoft on its integration into Copilot Checkout. The architecture is being built now.
The technology is moving faster than most people realize. Today, 23% of companies are already using agentic AI at least moderately — but within two years, that share is expected to rise to nearly 3 in 4 companies (74%), according to Deloitte’s survey State of AI in the Enterprise. On the business side, analysts project that agentic AI will drive $17.5 trillion in commerce by 2030, equivalent to 30% of all digital transaction value.
Consumer confidence is not keeping the same pace, though. Deloitte estimates that 70% of consumers are interested in exploring agentic AI in at least one area of payments, yet only 20% feel comfortable with it compared to other types of AI they already use. The reasons are specific: 58% cite security, data privacy, and hacking as top concerns, and 57% fear the agent making poor decisions, errors, or unauthorized actions on their behalf.
Those numbers are worth interrogating. Sharing personal data has become routine, and consumers accept broad data collection in exchange for discounts, personalized recommendations, or simply access to a service. However, the question shifts when agentic commerce eventually asks for your credit card information to book that Amalfi Coast trip, or your health records to schedule an appointment with your doctor. The technology may be seamless. The question is also whether your comfort with it will be too.
But comfort is only half the problem. As commerce enters this new era of delegation to autonomous systems, even a willing consumer runs into legal and operational questions the law has not yet answered: With a single click, does the consumer meaningfully understand (and therefore validly consent to) everything the agent is about to do? Who bears liability when an AI agent makes a mistake? How should markets verify an agent's authority to act? And can existing legal frameworks meaningfully govern transactions initiated not by humans, but by software acting on their behalf?
The legal infrastructure has not kept pace with the technology. That gap is what this series examines.
Continue reading · Part 2 of 4
The Rise of Agentic Commerce →New essays in your inbox, roughly monthly.
Legal analysis on the AI and privacy regulations that actually affect technology businesses. No noise, no boilerplate.
This article is for informational purposes only and does not constitute legal advice. Reading this essay does not create an attorney-client relationship.
