Amazon Blocks Meta Muse AI Agent Over Security Concerns

Starting September 20, Amazon.com, Inc. blocked Meta Platforms Incorporated's Muse AI agent from navigating and purchasing products on its retail site. Amazon alleges that Meta launched the agent without advance notice and designed it to browse without identifying as automated software. The retailer further claims Muse captures and stores customer credentials, creating serious privacy and security hazards and violating conditions of use regarding merchant consent.

Meta launched Muse on September 8 as a general-purpose assistant capable of multistep tasks, including financial management. Meta has rejected claims of mishandling sensitive information, stating that Muse uses protected storage for login information and cannot see passwords or payment details. This dispute occurs despite a commercial partnership where Meta uses Amazon's Graviton server chips and Amazon allows direct purchases within Facebook and Instagram feeds.

In response to the block, Meta CEO Mark Zuckerberg announced a partnership with Shopify to integrate Shop Pay checkout capabilities into Muse. This collaboration enables agentic checkout across Shopify-powered stores, allowing the AI to assist users through the shopping and payment process. While Zuckerberg and Shopify CEO Tobi Lutke frame the move as a way to simplify discovery, the conflict highlights a broader industry struggle over the ownership of the shopper relationship and the impact of AI agents on direct platform traffic.