Amazon Blocks Meta’s Muse AI Over Unauthorized Access

Amazon confirmed the move follows Meta’s failure to notify the retailer about the agent’s integration. An Amazon spokesperson stated that third-party services attempting to facilitate purchases on behalf of users must operate transparently and respect platform policies. The retailer specifically highlighted security risks, pointing to the agent’s failure to identify itself while browsing and the alleged capture of account information.

Meta previously claimed the agent could not access secure login or payment details, yet testing revealed the tool successfully completed transactions. The software has already faced scrutiny for reportedly viewing user message contents without explicit permission. This defensive stance aligns with Amazon’s broader strategy to control its ecosystem against AI-driven search and shopping tools. The company previously pursued legal action against Perplexity to limit its access to internal shopping data, though a judge ruled in favor of the startup this August. To further tighten security, Amazon has begun stripping item names and product images from confirmation emails, specifically targeting the data streams used by external AI scrapers.