Late Sunday evening, people attempting to purchase items on Amazon through Meta's AI assistant, Muse, began encountering an unusual error notification. According to GeekWire, the message stated: "Continued access by an unauthorized AI agent violates Amazon's Conditions of Use, to which our customers have agreed."

Put simply, Muse is not permitted to shop on the platform — and anyone hoping to make purchases on Amazon via the AI agent will need to look elsewhere.

It's tempting to interpret this restriction as a standoff between two tech powerhouses. After all, Amazon possesses its own collection of foundation models, not to mention one of the internet's most widely used inference platforms. With no legal requirement forcing them to grant Muse access, why would they?

EMBED_PLACEHOLDER_0

Yet there are also practical considerations that might explain why Amazon isn't eager to enter the agentic commerce space right now. Should Muse place an incorrect order, Amazon would bear the responsibility of resolving the situation — managing both the dissatisfied customer and the frustrated seller. While Muse has one of the lower hallucination rates among AI models, it remains considerably above zero. Even if Amazon recognizes potential in agent-driven commerce, they might prefer to wait through several more release cycles before launching anything.