Agentic commerce shifts commerce from people operating storefronts and checkouts to AI agents acting within a buyer’s delegated intent.
Agentic commerce is a model in which an AI system discovers products or services, evaluates alternatives, recommends options, and may complete a transaction for a consumer or business.
[ne4xtk]
[v6x6tl]
It applies when a customer interacts with an AI interface rather than navigating a conventional storefront, with the agent potentially handling search, comparison, negotiation, authorization, payment, and post-purchase tasks.
[ne4xtk]
[v6x6tl]
[9902c7]
The central change is that the agent becomes an active participant in the commercial process, not merely a passive search or customer-service interface.
[b615sw]
[9902c7]
flowchart LR
A["Human intent"] --> B["AI agent"]
B --> C["Discover products"]
C --> D["Evaluate options"]
D --> E["Obtain authorization"]
E --> F["Complete transaction"]
F --> G["Manage post purchase tasks"]
Expanded Chart
flowchart LR
A["Human intent"] --> B["AI agent"]
B --> C["Discover products"]
C --> D["Evaluate options"]
D --> E["Obtain authorization"]
E --> F["Complete transaction"]
F --> G["Manage post purchase tasks"]
Uses in Context
Conversational shopping: The term describes commerce conducted through an AI chat interface in which the agent finds, compares, and recommends products based on a user’s request.
[ne4xtk]
[v6x6tl]
Delegated purchasing: It refers to an agent identifying and buying within parameters established by the user, potentially without direct human intervention at checkout.
[9902c7]
[0ul53p]
Business procurement: Companies use the concept for agents that research suppliers, compare terms, and execute purchasing tasks on behalf of an organization.
[v6x6tl]
[9902c7]
Machine-readable commerce: The phrase increasingly denotes the infrastructure—product data, identity, authorization, payment, and checkout protocols—that allows agents to transact with merchants.
[xggwo8]
[xfi1r7]
Retail strategy: Merchants invoke agentic commerce to describe adapting catalogs, checkout flows, and customer relationships for AI-mediated discovery rather than only human browsing.
[xggwo8]
[b615sw]
History of Use
Origins
The available sources describe agentic commerce as an emerging industry term rather than attributing its coinage to a single academic paper or founder. Morgan Stanley defines it broadly as e-commerce “initiated or shaped by an AI system” after a consumer interacts with an agent, while SAP describes autonomous or semi-autonomous agents helping discover, evaluate, recommend, and complete commerce tasks.
[ne4xtk]
[v6x6tl]
The term grew out of the convergence of conversational AI, autonomous agents, digital commerce, and machine-readable payment systems rather than from one documented invention.
[v6x6tl]
[xggwo8]
Evolution
Early formulation — AI-assisted discovery: The concept initially emphasized agents helping consumers find and compare products, extending conversational search into shopping.
[ne4xtk]
[v6x6tl]
Expansion — delegated transactions: It broadened to include recommendations, authorization, negotiation, checkout, and payment carried out on a customer’s behalf.
[9902c7]
[f8fxry]
Infrastructure phase — protocols and agent identity: By the mid-2020s, attention shifted toward common standards for product discovery, agent-to-agent communication, identity, authorization, and payment; industry commentary compared the need for a common language to HTTP’s role on the web.
[xggwo8]
Best Real-World Examples
Klarna Shopping Search — Klarna — an AI shopping-search experience inside ChatGPT that surfaces real-time prices and merchant listings across a large product catalog.
[fd3nsk]
Rye — Rye — a startup example of infrastructure designed to enable universal checkout across fragmented merchants and storefronts.
[xggwo8]
Skyfire — an agent-native identity and payments company building primitives for AI agents that need to transact.
[xggwo8]
Nekuda — a startup focused on identity and payment infrastructure for agentic commerce.
[xggwo8]
Catena Labs — an emerging infrastructure provider working on foundational components for autonomous transactions.
[xggwo8]
Universal Commerce Protocol — a commerce standard promoted by Google and retail partners to support an end-to-end shopping journey from discovery through post-purchase service.
[xggwo8]
Case Studies
Klarna and conversational product discovery. Klarna introduced an AI-powered shopping-search application inside Chat GPT that uses a Product Search MCP server to surface current prices and merchant listings.
[fd3nsk]
The model changes the entry point to shopping: rather than beginning with a retailer’s homepage or a conventional search-results page, the buyer begins with an intent expressed in natural language.
[ne4xtk]
[fd3nsk]
This illustrates the discovery side of agentic commerce, where the agent filters a broad market and presents options before the user reaches a conventional checkout.
Startup infrastructure versus incumbent storefronts. Smaller companies such as Rye, Skyfire, Nekuda, and Catena Labs are targeting gaps that arise when autonomous agents must identify themselves, obtain permission, access merchant systems, and pay securely.
[xggwo8]
Industry coverage characterizes these startups as building universal checkout, agent-native identity, and payment primitives while larger firms concentrate on protocols and distribution.
[xggwo8]
The case demonstrates that agentic commerce is not simply an AI shopping feature; it is also an infrastructure layer in which startups can address interoperability problems before they are absorbed into mainstream platforms.
Protocol competition and standardization. The emerging ecosystem includes OpenAI and Stripe’s Agentic Commerce Protocol for checkout, Google’s Universal Commerce Protocol for a broader shopping journey, and payment-oriented systems associated with Mastercard, Visa, and other networks.
[xggwo8]
[0ul53p]
The competing approaches reflect different definitions of the problem: some prioritize immediate checkout, while others address catalog discovery, post-purchase interactions, agent-to-agent communication, or payment authorization.
[xggwo8]
This shows that the field remains unsettled: agentic commerce is evolving from a product experience into a standards and governance contest over how agents discover offers, represent user intent, prove authority, and execute transactions.
