Personal AI agents like Meta’s Muse, Instinct, and ChatGPT’s Dots are leading the charge into a new era of consumer AI that does more than just answer questions. These agents can handle tasks such as booking flights, making dinner reservations, and ordering groceries. However, users often run into roadblocks when the websites they visit prevent these agents from completing their tasks.
Recently, Amazon blocked Meta’s Muse from accessing its retail site, stopping the AI from browsing or purchasing items from its catalog. This problem isn’t limited to Amazon; users have reported similar issues across various platforms. Some blocks are intentional, while others arise from standard anti-bot measures that websites use to guard against spam and malicious behavior.
This situation leaves consumers confused, as they often can’t tell whether a block is intentional or accidental, which leads to frustration with both the retailer and the AI agent trying to complete the transaction. For instance, complaints have surfaced about Muse’s inability to finalize purchases on Walmart’s site. However, a Walmart spokesperson said these blocks were not intentional – Walmart is actually partnered with Muse, as announced at Meta’s Connect developer conference in September, and the company said it wants to be available through AI agent experiences.
The challenge intensifies when websites implement human verification processes, such as requiring users to click a button to confirm they are human. If the interaction is interrupted, the verification can fail, resulting in the AI agent being blocked.
To address this challenge, industry partners including Meta, Walmart, Stripe, Sierra, Genesys, Rocket, NiCE, and Decagon are working on an open standard. This protocol aims to define how AI agents can effectively communicate with businesses, distinguishing between legitimate bots acting on behalf of users and harmful bots. Meta emphasized that this will focus on agent-to-agent communication specifically for online commerce.
Consumer confusion is widespread, with reports indicating that while some companies welcome AI agents, others are resistant. For example, Delta Airlines has stated that it does not currently have a partnership or integration that enables a third-party AI agent to shop for or book Delta flights on a customer’s behalf on its digital platforms. Heena Chavda, Delta’s General Manager of Global Communications, mentioned that any future integration would prioritize security and customer experience and that they are evaluating how technology, including AI agents, could enhance customer engagement.
United Airlines pointed to its Terms of Use policy, which says users agree “not to use any robot, spider, other automatic device, or manual process to monitor or copy this website” without United’s prior written permission. United didn’t respond to a follow-up question about whether it is blocking specific personal AI agents like Muse. Various brands, including Yelp, eBay, and several airlines, have been identified by early adopters as blocking AI agents. For instance, Yelp only allows agents to access its content if they have paid for data licensing, meaning that many attempts to use its services could fail without a formal agreement.
A couple of users reported that eBay suspended their accounts because of their use of agentic AI.
Primary Issues Faced by AI Agents
1.
Cloudflare’s role in blocking AI agents
One major company involved in this space is Cloudflare, which offers sites protection from AI agents that could be training on its data and managing other automated activity.
