In a significant move toward an automated e-commerce landscape, Salesforce has integrated a new protocol developed by Stripe and OpenAI, enabling AI “agents” to conduct purchases on behalf of consumers. This partnership allows businesses using Salesforce’s Commerce Cloud to accept payments processed by Stripe for transactions initiated by AI within conversational interfaces like ChatGPT. The collaboration marks the first time a major e-commerce platform has adopted the Agentic Commerce Protocol (ACP), an open-source standard designed to create a universal language between AI agents, online retailers, and payment systems.
The core of the initiative is to build the infrastructure for a future where shopping is delegated to autonomous software agents. Instead of manually adding items to a cart and completing checkout forms, consumers can instruct an AI to find and purchase products based on specified criteria. For merchants, this opens a new sales channel directly within popular AI platforms, potentially reaching millions of users where they are already seeking information and product recommendations. Research from Salesforce suggests a growing consumer readiness for this shift, with nearly half of shoppers who use AI indicating they would trust an AI agent to make a purchase for them.
A New Standard for AI-Driven Transactions
The Agentic Commerce Protocol, co-developed by Stripe and OpenAI, serves as a foundational layer for this new form of e-commerce. It functions as an open-source rulebook, similar to how HTML provides a standard for web browsers, ensuring that different AI agents and retail systems can communicate without requiring custom integrations for each platform. This standardization is crucial for scalability, as it prevents a fragmented ecosystem where businesses would need to build separate connections for every AI assistant their customers might use. The protocol defines how retailers can present their product catalogs and checkout systems to AI agents in a machine-readable format.
Security is a central component of the ACP framework. Transactions are processed using secure tokens, such as Stripe’s Shared Payment Token, which allows payment information to be transmitted without exposing raw credit card details to the AI agent. The AI facilitates the purchase, but the merchant remains the merchant of record, retaining control over whether to accept or decline the transaction based on payment and fraud signals. This architecture ensures that businesses maintain their direct relationship with the customer, handling all post-purchase experiences while leveraging the AI as a new discovery and sales channel.
Integration with Salesforce and Stripe Infrastructure
For businesses already using Salesforce’s Commerce Cloud, the integration is designed to be seamless. Transactions originating from AI agents will flow through the same payment infrastructure as conventional online purchases, leveraging the existing Salesforce Payments system, which is built on Stripe’s platform. This means merchants can enable AI-driven sales without overhauling their current payment and order management workflows. Nitin Mangtani, General Manager of Commerce Cloud and Retail at Salesforce, described the collaboration as delivering a “unified system designed for the future of agentic commerce,” aimed at creating a faster and more personalized path to purchase.
The system supports a variety of payment methods through Stripe, including its one-click checkout product, Link, which stores customer payment and shipping details to expedite transactions. By connecting with Salesforce’s Customer 360 platform, retailers can deploy AI agents on their own websites, combining customer data from marketing, sales, and service to inform the automated shopping experience. This holistic view allows the AI to make more relevant and personalized recommendations, ultimately driving revenue and fostering deeper customer loyalty.
The Emerging Landscape of Conversational Commerce
This partnership signals a broader industry bet on the rise of “agentic commerce,” where AI transitions from a background assistant for product recommendations to an active participant in the transaction itself. The initial implementation of the Agentic Commerce Protocol is powering an “Instant Checkout” feature within ChatGPT, allowing users in the United States to buy products from participating merchants, such as those on Etsy and soon Shopify, without leaving the chat interface. This model turns AI platforms into new storefronts, enabling businesses to capture sales at the moment of discovery.
While the technology is now in place, questions remain about the pace and scale of consumer adoption. The success of AI-mediated commerce will depend on whether shoppers embrace making purchases through chatbots rather than visually browsing a retailer’s website. However, proponents argue that the convenience of delegating complex shopping tasks to an AI—such as finding a specific product that meets multiple criteria—will be a compelling value proposition. As Maia Josebachvili, Chief Revenue Officer of AI at Stripe, noted, the collaboration aims to help merchants “thrive in the agentic commerce era” by turning discovery directly into purchase.
Future Outlook and Industry Implications
The partnership between Salesforce, Stripe, and OpenAI represents a significant step in building the technical foundation for a new era of e-commerce. The open-source nature of the Agentic Commerce Protocol encourages broad adoption, allowing it to work across different payment processors and AI agents beyond just Stripe and ChatGPT. This flexibility is intended to foster a competitive and innovative ecosystem, preventing any single company from dominating the emerging market for AI-driven shopping.
As more businesses connect to the protocol, the capabilities of AI shopping agents are expected to grow. Future developments will likely focus on enhancing the personalization of these agents, improving their ability to understand complex user requests, and expanding the range of products and services they can purchase. For the broader retail industry, this shift necessitates a reevaluation of online strategy, moving beyond traditional websites and mobile apps to consider how their products and brand are represented within conversational AI platforms. The companies that adapt to this machine-to-machine interaction model are poised to gain a significant advantage in the evolving digital marketplace.