AI agents are becoming capable enough to shop, book flights and make reservations, but the web infrastructure they depend on is not yet designed to trust them.
The supplied source says users of personal agents are increasingly hitting website blocks. Some restrictions are intentional, while others are side effects of anti-bot systems that cannot distinguish a legitimate agent acting for a customer from scraping or malicious automation.
Amazon has intentionally blocked Meta’s Muse agent from its retail site, according to the source. Walmart, by contrast, says it wants to support agents but has seen automated sessions disrupted by human-verification checks.
The problem extends into travel. Delta said it does not currently support third-party AI agents booking flights on its digital platforms, while United’s terms prohibit unauthorized automated access.
This creates a new infrastructure layer for agentic commerce: identity and permission.
Meta, Walmart, Stripe, Sierra, Genesys, Rocket, NiCE and Decagon are working on an open standard for agent-to-business communication. The goal is to help businesses distinguish trusted agents from harmful bots.
For investors, the issue is larger than consumer convenience. If agents become a meaningful commerce channel, merchants that block them may lose transactions, while companies that control agent identity, payments and permissions could capture new economic value.
BTI’s bottom line: the next bottleneck for consumer AI is not model intelligence. It is access. Agentic commerce needs a trust standard that lets websites know when an automated visitor is acting with a real customer’s permission.