Personal AI agents such as Meta’s Muse, Instinct, and ChatGPT’s Dots can now book flights, reserve tables, and order groceries without requiring users to install complex software. However, these tools are frequently blocked by websites that refuse access or fail to verify their presence.
Amazon recently stopped Muse from browsing its retail site or making purchases. Similar complaints have surfaced across social media, where users report their agents being rejected by various platforms. Some blocks are intentional, while others stem from standard anti-bot measures designed to stop spam.
Users often cannot tell if a block is accidental or deliberate, leaving them frustrated with both the retailer and the agent. TechCrunch has seen reports that Muse failed to complete a purchase on Walmart’s site. NBC also noted this issue during Walmart’s testing phase.
A Walmart spokesperson stated these failures were not intentional. The company announced a partnership with Muse at Meta’s Connect developer conference in September. Walmart aims to be available through AI agent experiences alongside traditional channels.
The core issue appears to be a verification step. The website presents a button requiring a click to prove the visitor is human. If this process is interrupted, the check fails and the agent is removed.
Current methods for verifying humans and stopping bots may not fit the agent-first era of the web. To fix this, industry partners including Meta, Walmart, Stripe, Sierra, Genesys, Rocket, NiCE, and Decagon are developing an open standard. This protocol will define how AI agents communicate with businesses to separate legitimate user agents from malicious ones.
Consumer confusion remains high while companies adopt different approaches. While Walmart is working towards compatibility, many others are not.
Flight booking is a primary use case, yet airlines often reject agent requests. Delta stated it is taking steps to protect customers from unauthorised automated activity. The company noted that opening up to agents requires prioritising security and the customer experience.
“While Delta 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 our digital platforms, we’re continuing to evaluate how new technology, including external AI agents, could enhance the way customers engage with Delta,” said Heena Chavda, General Manager of Global Communications at Delta.
United Airlines was less direct. A spokesperson pointed to its Terms of Use policy, which prohibits using robots, spiders, or automatic devices to monitor or copy the website without prior written permission. The company did not answer a follow-up question about blocking specific agents like Muse.
Complaints on social media reference brands blocking agents, including Yelp, eBay, Zillow, Pizza Hut, Adidas, and various airlines. Some users claim blocking activity has increased recently.
Two individuals reported eBay suspending their accounts due to agentic AI use. Another said Google removed Instinct while it cleaned up a Gmail inbox.
Yelp stated it does not permit non-human traffic unless the agent has paid to access data through its licensing program. Agents trying to get quotes, join waitlists, or book tables will likely fail without a formal partnership.
eBay told TechCrunch it does not prohibit all third-party shopping agents. Its policies restrict unauthorised actions such as automated scraping and model training.
Adidas, Zillow, and Pizza Hut did not respond or declined to comment by the time of publication.
Cloudflare, a major company in this space, offers protection against AI agents training on its data. Some suspect Cloudflare and similar content delivery networks are disrupting Muse traffic. A change to crawler defaults on September 15 allows site owners to block AI training while permitting other bots. Sites previously set to block AI bots for security reasons were shifted to block AI agents on any pages containing ads.
When asked about the blocking of personal AI agents, Cloudflare said it did not have specific data to share. It pointed to its public data hub, Cloudflare Radar. The site shows an increase in bot activity but does not differentiate between good and bad bots.
Cloudflare also runs a marketplace where AI bots pay for data access. The company recently began testing a new web browser built specifically for AI agents.
Meta’s decision to form specific brand partnerships for Muse access and focus on new standards seems the right approach. This assures customers that the agent is welcome at partner sites, making access issues reportable bugs.
Meta maintains a list of connectors inside the Muse app, which continues to grow. Many newly announced partners have not yet appeared on this list. Given Muse’s rapid expansion, including tools for small businesses, a bigger reorganization of the connector list is likely soon.
“Turning away a personal agent means turning away the customer behind it. We’d rather work with businesses to address the underlying concerns than see them lose that customer,” Meta said in its announcement about developing a standard. “There is a path that is good for users and good for businesses. It starts with clear norms, gives businesses more predictable control, and evolves toward a more efficient way for agents to work together.”
What it means
Until a universal standard is adopted, users must expect friction when trying to use AI agents for tasks like booking travel or shopping. Success depends on whether a specific retailer has agreed to work with the agent’s provider.
