AI Agents Face New Gatekeepers as Websites Block Automated Access
The rise of personal AI agents capable of performing tasks on users' behalf is met with significant challenges as websites frequently block their activities, causing user frustration. While some blocks are intentional, others stem from traditional anti-bot measures, prompting industry leaders to develop an open standard for agent-to-business communication. This initiative, alongside Meta's strategy of brand partnerships, aims to streamline online commerce for AI agents and enhance customer experience.
A new era of consumer AI is emerging as personal agents such as Meta’s Muse, Instinct, and OpenAI’s Dots move beyond answering questions to performing tasks on users’ behalf, including shopping, booking flights, and making reservations.
But as these agents become more capable, they are increasingly running into digital barriers on websites, creating friction for users who expect them to complete tasks without human intervention. Amazon recently blocked Meta’s Muse from browsing its retail site or making purchases, while other reported disruptions appear to result from conventional anti-bot systems designed to prevent spam, scraping and malicious activity.
The problem is not limited to Amazon, with users reporting difficulties involving Walmart, airlines, Yelp, eBay, Zillow, Pizza Hut and Adidas. Walmart, despite being a Muse partner, said problems affecting some users were not intentional and may stem from its human-verification system, while Delta said it is evaluating external AI agents with security and customer experience in mind; United’s terms also restrict unauthorised automated activity.
Yelp does not permit non-human traffic unless agents have formal data-access arrangements, while eBay said it does not prohibit all third-party shopping agents but restricts unauthorised agents and activities such as automated scraping and model training.
The conflict also extends to infrastructure providers such as Cloudflare, whose updated AI-bot policies introduced on September 15, 2026 allow site owners to distinguish between search, training and agent activity, including blocking agents on advertising-supported pages. The emerging problem is therefore bigger than whether individual companies want to welcome AI agents: the web’s existing security and verification systems were largely built around distinguishing humans from bots, not distinguishing legitimate agents acting for users from malicious automation.
In response, Meta, Walmart, Stripe, Sierra, Genesys, Rocket, NiCE and Decagon are working on an open standard for agent-to-agent communication in online commerce, potentially giving businesses a clearer way to recognise authorised AI agents while allowing users to know where their digital assistants are actually welcome.