eBay Banned AI Shopping Agents; Opened the Door to Approved Bots

eBay announced that as of February 2026 it will ban artificial intelligence–powered shopping agents from operating on the platform.

eBay’s decision was taken at a time when, with the rapid spread of AI-based commerce tools in the retail sector, marketplaces are reassessing the impact of automated buyers on human customers and existing business models. This policy change was announced via an email sent to eBay users. The platform had previously taken technical steps aimed at limiting automated access.

New Rules Against Autonomous Shopping Bots

eBay stated that as of 20 February 2026, chatbots, large language model–based agents and other automated systems will be explicitly prohibited from placing bids or completing purchase transactions on the auction platform unless prior permission is obtained. The updated user agreement clearly defined that “automated means” used to access services cannot be used without explicit approval from eBay.

The update also expanded the scope of existing provisions that restrict robots, scrapers and other automated tools from accessing eBay services. In addition, with changes made to the platform’s robots.txt file, it was emphasized that checkout processes are reserved exclusively for human users.

Although the robots.txt protocol cannot technically prevent all bot access, the updated user agreement provided eBay with a legal basis against unauthorized AI activities. The company stated that it took this step in order to protect the integrity of the marketplace and to maintain a fair competitive environment for human buyers.

Pressure From Agentic Commerce Increased

The ban decision came at a time when new shopping models referred to as agentic commerce have begun to be rapidly tested in the technology sector. These models envision consumers delegating product discovery and purchasing decisions to software agents instead of directly visiting websites.

Conversational AI platforms such as ChatGPT have begun to integrate early-stage commerce features that enable direct purchases from certain e-commerce sites. Similarly, Perplexity offers one-click checkout options for paid users, while Google is working on a Universal Commerce Protocol aimed at standardizing how shopping agents interact with retailers.

Amazon is also testing features that allow users to purchase products from external brands directly within its own application. However, it has also been reported that Amazon has taken legal steps to prevent unauthorized AI agents from operating on its own marketplace. This situation points to growing concerns across the industry regarding uncontrolled automation.

eBay Takes a Cautious Approach

While eBay’s updated user agreement introduced a clear ban, it also pointed to a more selective future. By stating that certain AI-based entities may operate on the platform if prior approval is obtained, the company implied that tightly controlled agents could be allowed in the future.

eBay CEO Jamie Iannone had previously indicated that the company could participate in OpenAI’s Instant Checkout program. Iannone also stated that eBay is experimenting with its own agentic shopping experiences.

This dual approach showed that eBay is not completely rejecting AI-powered commerce, but instead wants to strictly control how and when it is deployed. Experts noted that this cautious stance is largely rooted in eBay’s auction-based business model, as automated bidding systems carry the risk of disrupting price dynamics.

Economic and Operational Concerns

One of eBay’s main concerns was economic. The platform charges a variable “final value fee” on sales and earns more revenue from higher-priced transactions. Autonomous agents focused on securing the lowest possible price could win auctions at lower amounts, leading to reduced selling prices and potential revenue loss for eBay.

There are also operational risks. Fully automated agents could generate heavy traffic on servers, straining infrastructure. In addition, since such agents would not respond to cross-selling or promotional efforts aimed at human users, seller engagement and advertising effectiveness could decline.

Trust and safety issues also came to the forefront. For eBay, which continues to combat counterfeit goods and unsafe sales, monitoring autonomous purchasing systems and assigning responsibility could become more complex.

Industry-Wide Debate

The decision became part of a broader debate on how agentic artificial intelligence should be integrated into e-commerce. Proponents foresee a future in which AI agents anticipate consumer needs, evaluate options, negotiate prices and complete transactions independently. Consulting firms such as McKinsey are already highlighting such multi-step, autonomous shopping scenarios.

Critics, however, argue that this approach could reduce direct consumer interaction with websites, weakening brand visibility, product discovery and upselling opportunities. In this case, retailers may be forced to develop new machine-to-machine systems alongside interfaces designed for human users.

A Controlled Path Forward

By banning unauthorized AI shopping agents while not fully closing the door to approved uses, eBay adopted a controlled agentic commerce approach. The company acknowledged that automation could play a role in the future, but made it clear that this should occur under conditions that protect marketplace fairness, revenue balance and user trust.

As AI-powered shopping tools continue to evolve, eBay’s decision stood out as an important example of how large marketplaces are attempting to strike a delicate balance between innovation and control.

eBay Launches SpeedPAK Service to Simplify International Shipping for German Sellers