The next hurdle for AI agents: getting websites to let them in
Consumer AI agents can book flights, reserve tables and order groceries for users, but websites increasingly block them, preventing tasks from being completed. The problem matters because existing anti-bot protections can treat helpful agents like malicious traffic, leaving customers unsure whether a retailer is deliberately blocking access or a security check has simply failed.
Amazon has blocked Meta’s Muse from browsing and buying products, while Walmart says reported failures on its site were unintended and may stem from a human-verification button that agents cannot complete. Meta, Walmart, Stripe and other companies are developing an open standard for agents to communicate with businesses during online commerce. Meanwhile, reports of airline sites rejecting agents point to wider uncertainty about how services will handle this emerging way of shopping and booking.
- Websites are blocking AI agents from completing everyday tasks.
- Walmart says its reported blocks were accidental.
- Industry partners are working on an open commerce standard.
New here? Start with this
AI agents are software tools that perform tasks on the internet on behalf of users. They can book flights, reserve restaurant tables and order groceries without human intervention. This technology promises to save consumers time by automating routine online shopping and booking tasks.
However, many websites treat these agents like malicious bots and block them from accessing their platforms. Existing security systems designed to prevent cybercriminals from attacking websites cannot always distinguish between helpful agents and harmful traffic. This creates confusion for customers, who may not know whether a retailer is deliberately blocking access or if a security check has simply failed.
Major technology companies including Amazon, Meta, Walmart and Stripe are working to resolve this by developing an open standard for agents to communicate with businesses. Reports indicate that Amazon has blocked Meta's Muse from browsing and buying products, whilst airline sites have also rejected agents. Until websites agree on how to handle these tools, customers may find their automated tasks failing without clear explanation.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
AI agents represent a natural progression in how consumers interact with online services, allowing users to complete legitimate purchases and bookings through automated assistants that act on their behalf. Blocking these agents arbitrarily treats beneficial software the same as malicious bots, preventing customers from accessing services they have every right to use whilst creating confusion about whether a transaction failure is technical or deliberate. An open standard would clarify legitimate agent use, enhance accessibility, and drive commerce forward without preventing necessary security measures.
The case against
Website operators have legitimate reasons to restrict automated access to their platforms, including protecting against fraud, preventing data scraping, and maintaining control over how their services are used and presented to customers. Agents accessing sites without explicit permission may violate terms of service and create security vulnerabilities that threaten customer data and platform integrity. Companies should retain autonomy over their own systems and business models rather than being forced to accommodate any bot claiming to represent a consumer.