The landscape of digital commerce is undergoing a tectonic shift. We are moving beyond the era of the “point-and-click” shopping experience toward an age of “agentic commerce,” where autonomous AI assistants navigate the web, select products, and complete financial transactions on behalf of their users. With the meteoric rise of Meta’s Muse—which amassed five million downloads in just 22 days—and OpenAI’s development of its “Dots” agent and a proprietary “ChatGPT Wallet,” the race to capture the AI-powered consumer is officially underway. However, as the technology accelerates, a critical question remains: are these agents a genuine evolution of retail, or are they a high-stakes experiment building on shaky foundations? The Main Facts: The New Wave of Autonomous Purchasing The core promise of these new AI agents is frictionless utility. Unlike a static chatbot that merely provides product recommendations, agents like Muse and the forthcoming Dots are designed to understand user intent and execute complex workflows. Muse has set the current industry benchmark, integrating Stripe’s Link payment infrastructure. This allows the AI to facilitate purchases directly, utilizing one-time-use card functionality to enhance security. Plans are already in motion to integrate Shop Pay, further embedding the agent into the existing merchant ecosystem. OpenAI is not standing still. Its rival project, “Dots,” aims to replicate this functionality, with reports suggesting that the company is actively developing a “ChatGPT Wallet.” This would allow users to store multiple payment credentials directly within the OpenAI ecosystem, enabling the AI to act as a persistent, authenticated buyer across various online storefronts. A Chronology of AI Commerce: From Failure to Rebirth The current optimism surrounding AI agents belies a history of trial and error. Early 2024: OpenAI launched its “Instant Checkout” feature, an early attempt to integrate direct purchasing within ChatGPT. The goal was to remove the friction of navigating traditional merchant websites. March 2024: OpenAI quietly shuttered the Instant Checkout initiative. The company cited a lack of “the level of flexibility that we aspire to provide,” marking a significant retreat from direct, platform-owned payment processing. Late 2024 (Present Day): The industry has pivoted. Instead of attempting to replace the retailer’s checkout journey, the new generation of agents acts as a mediator. They perform the “heavy lifting” of shopping—finding items and initiating the process—while still utilizing the merchant’s existing payment rails. This shift in philosophy is what analysts believe has sparked the current wave of adoption. Supporting Data and the Scalability Question The numbers behind Muse’s launch are staggering. Reaching five million users in 22 days highlights a massive consumer appetite for automation. However, industry experts caution against conflating adoption speed with commercial viability. The technical integration of these agents into existing payment gateways like Stripe is a triumph of software engineering, but it represents a “short-term technical fix.” The challenge, according to PSE Consulting’s Managing Director Chris Jobes, is that the current infrastructure was never designed for non-human actors. Current merchant payment rails rely on a chain of trust between the bank, the merchant, and the known human user. By inserting an AI agent into the middle, that chain becomes significantly more complex. If an agent commits a fraudulent transaction, or if an item fails to arrive, the liability is currently poorly defined. The existing framework is built on human identity; once that identity is abstracted by an AI proxy, the “plumbing” of the financial system begins to leak. Official Responses and Expert Analysis The industry is currently divided between tech giants eager to capture market share and risk consultants concerned about the lack of guardrails. Chris Jobes of PSE Consulting has emerged as a leading voice of caution. His critique centers on the concept of the “agent as a wallet.” “Giving agents access to consumers’ payment credentials makes buying smoother, but it also concentrates risk around identity, authority, and fraud,” Jobes notes. “Merchants have less visibility of who is behind the purchase. When goods fail to arrive or a payment is disputed, more of that risk shifts to the wallet layer. Furthermore, an agent that can pay is a prime target for fraud. If criminals can impersonate an agent, manipulate its instructions, or exploit weaknesses in how it proves its authority, they could trigger purchases the customer never properly approved.” Beyond security, there is the issue of influence. If an AI agent is funded by advertising revenue, can it ever be truly impartial? If a user asks an agent to find “the best running shoes,” but the agent is incentivized to prioritize brands that have paid for preferential placement, the trust dynamic between the user and their AI assistant is irreparably damaged. Implications: The Need for New Rules The rise of agentic commerce creates a unique set of implications for the retail sector, ranging from regulatory hurdles to a fundamental change in the merchant-customer relationship. 1. The Regulatory Vacuum There is currently no unified regulatory framework for agentic commerce. Who is responsible when an AI makes a mistake? If an agent purchases a high-value item without the user’s explicit confirmation, is it the user’s fault for delegating, the agent provider’s fault for poor logic, or the merchant’s fault for accepting the payment? Until standards for identifying agents and verifying their authority are established, the sector will struggle to achieve mainstream, large-scale stability. 2. The Erosion of Brand Loyalty When an agent handles the purchasing process, the consumer’s relationship with the brand changes. If the AI is making the choice based on price, availability, or an algorithm-driven recommendation, the retailer loses the opportunity to build brand affinity through their own website design and marketing efforts. Merchants are effectively being relegated to “fulfillment centers” for AI agents. 3. The Trust Paradox For AI shopping to scale, transparency is non-negotiable. Consumers must have absolute clarity on why an agent chose a specific product. If the recommendation engine is a “black box” influenced by back-end advertising deals, consumers will eventually view these agents as glorified telemarketers rather than helpful assistants. 4. The Future of Authentication The industry must move toward common standards for “machine-to-machine” authentication. Just as we have SSL certificates for websites, we may soon require “agent credentials” that can be verified by merchant payment gateways in real-time, confirming that the agent is acting within the specific, authorized scope defined by the human user. Conclusion: A Work in Progress The arrival of Muse and the impending launch of ChatGPT’s wallet represent a significant technological milestone. We have successfully automated the “checkout journey.” However, as Jobes wisely points out, solving the technical checkout is not the same as solving the commercial framework. For now, agentic commerce is in its “wild west” phase. It is a period of rapid innovation, high enthusiasm, and significant risk. The agents are ready, but the rules are not. Until the industry can agree on the plumbing—the protocols for identity, the protocols for dispute resolution, and the protocols for transparency—this technology will remain a high-tech convenience rather than the foundation of a new global retail economy. The next twelve months will be critical. If the major players (Meta, OpenAI, and their financial partners) move toward establishing common standards, we may see a rapid maturation of the market. If, however, they continue to push for proprietary, closed-loop systems that prioritize their own growth over systemic safety, we should expect a bumpy road ahead, potentially marked by the same “quiet retreats” that characterized the end of OpenAI’s earlier forays into commerce. Ultimately, the goal of an AI assistant should be to serve the human user. If the price of that service is a loss of financial security or the erosion of consumer trust, the revolution will be short-lived. The technology is here, but the framework is still under construction. Stay informed: Our editor carefully curates two newsletters a week filled with up-to-date news, analysis, and research. Click here to subscribe to the FREE newsletter sent straight to your inbox. Why not follow us on LinkedIn to receive the latest updates on our research and analysis? Post navigation Global Scale Meets Niche Precision: TD Synnex to Acquire BlueStar in Strategic Distribution Shift Levi Strauss & Co. Navigates Digital Transformation Amidst Shifting Consumer Trends