Websites increasingly block AI agents from accessing their services

Telescope log, October 17. Sources report a widening pattern of resistance across the open web toward autonomous AI agents acting on users’ behalf. According to The Verge, companies including OpenAI, Meta, Google, and xAI have each promoted agents capable of completing tasks such as booking reservations, managing inboxes, or purchasing goods without direct human input at every step.

Users of several such tools, named as Meta’s Muse, xAI’s Grok, OpenAI’s Dots, and a service called Instinct, have reported in recent weeks that their agents are being blocked when attempting to carry out tasks on major websites. Yelp is cited among the sites involved, though the source material available here does not specify the mechanism or stated reason for these blocks.

The report frames this as a structural tension rather than an isolated incident. Multiple large technology firms have invested in agent products as a central feature of their AI offerings, while, separately, the sites those agents are meant to operate on are moving to restrict their access.

No further technical detail, statement from the named companies, or site operators is available in the material provided. The Specola notes the matter as reported, without further comment, pending additional primary sourcing on the scope and methods of the blocking described.