Amazon says it blocked Meta's Muse AI assistant from accessing Amazon.com to shop on users' behalf, citing ToS violations, security risks, and merchant consent

Access to your Amazon shopping history is no longer available to Meta's Muse AI assistant.

Amazon has blocked Meta's Muse AI assistant from accessing Amazon.com to facilitate shopping on behalf of users. Amazon stated that Muse violated its Conditions of Use, posed security risks by potentially capturing user credentials, and operated without Amazon's consent. Meta's Muse agent reportedly does not identify itself when browsing Amazon's site and may store customer credentials. Amazon is taking this action despite having existing business partnerships with Meta, including a significant cloud computing deal. This move by Amazon follows previous efforts to block other AI shopping agents, such as Perplexity's Comet, from its platform.

This standoff highlights a critical tension in the burgeoning field of AI agents: control over the customer relationship and data in online commerce. Amazon's stance protects its lucrative advertising business, which relies on users browsing its site and seeing sponsored products, and its own developing AI shopping features. By blocking Meta, Amazon asserts its authority over its platform, setting a precedent for how AI agents must interact with e-commerce sites.

The dispute also underscores the challenges of interoperability and consent in agentic AI development. While Meta claims Muse securely stores credentials without direct visibility, Amazon views any unconsented access as a security risk and a violation of its terms. This conflict signals that companies will likely need explicit agreements and transparent operating protocols for AI agents to engage with third-party platforms, potentially slowing widespread adoption.