OpenAI, Meta, Google, and others are pushing AI agents that can book reservations, shop, or manage accounts on a user's behalf. But many of the web's most popular services are refusing to let those agents in. Over recent weeks, users of Meta's Muse, xAI's Grok, OpenAI's Dots, and Instinct have reported being blocked from completing tasks on major websites. Yelp says it only allows agents that pay for its data licensing program; Delta Air Lines is preventing ticket bookings; JPMorgan Chase and Charles Schwab are blocking account access over security concerns.
The standoff between Amazon and Perplexity shows how contentious this is becoming. Amazon banned Perplexity's shopping feature and sued the AI startup, arguing that third-party agents should respect service providers' decisions. Perplexity countered that an AI agent is an extension of the user and that Amazon simply wants to protect its ad-driven traffic. A judge initially ordered Perplexity to stop, but an appeals court overturned that ruling in August, and Amazon has since filed an amended complaint.
These blocks are already affecting how the AI industry's promises land with users. Chris Canal, CEO of AI safety evaluator EquiStamp, told The Verge that the value of companies like OpenAI and Anthropic is tied to agents being able to do things online—and that value drops when companies actively prevent it. Meta, for its part, says turning away a personal agent means turning away the customer behind it. It has partnered with Walmart, Gap, Sephora, and Best Buy to let Muse shop there, though even those sites sometimes trip over anti-bot systems. Infrastructure provider Cloudflare has also made it easier for sites to block AI crawlers, which may keep bots out but, as Canal notes, could make the web more annoying for everyone.