AI
Websites block AI agents, exposing a practical limit on the technology’s ambitions
AI agents marketed as digital assistants are encountering growing resistance from the websites they need to use. Yelp restricts agent access to licensed data users; Delta blocks agent-made bookings; JPMorgan Chase and Charles Schwab block account access; and Amazon remains in litigation with Perplexity over automated shopping. Some failures also come from anti-bot systems rather than explicit bans. The conflict threatens agents’ usefulness, while raising unresolved questions about site control, advertising economics, security and whether an agent should count as an extension of its user.
Editorial summary · Publisher article
What matters
- Yelp, Delta, JPMorgan Chase and Charles Schwab have imposed restrictions or blocks on AI-agent access.
- Amazon has pursued legal action against Perplexity’s shopping agent, though an appeals court overturned an earlier order against it.
- Meta is pursuing partnerships with retailers including Walmart, Gap, Sephora and Best Buy, while Shopify has opened its storefront network to agents.
- Anti-bot systems can block agents alongside other bots, potentially making ordinary web use more cumbersome through additional human checks.
A dispute over who controls online access
Website operators argue that automated services must respect their participation rules and protect security, infrastructure and business models. Agent companies counter that a personal agent acts on behalf of the customer. The disagreement is especially sharp where agents reduce human visits and therefore advertising exposure.
Compatibility is not guaranteed
Even willing sites may fail to support agents because fraud and scraping defenses cannot reliably distinguish them from malicious bots. That means agent usefulness depends not only on model capability but also on authentication, commercial agreements and web infrastructure.
Context & caveats
- Several reported blocks are based on user reports, and OpenAI and Google had not responded to the publication’s requests for comment.
Why this made the edition
The report supplies concrete examples of websites and financial services blocking AI agents, includes company responses and legal context, and explains how anti-bot systems create additional barriers. It is substantive and directly relevant to the practical limits of agentic AI.
Source material