Archive · 22 Sep 2026
Amazon Blocks Meta’s Muse Over Unauthorized Shopping Access
Muse is an AI helper made by Meta that can shop and complete other tasks for people.
Amazon stopped Muse from using its website to browse or buy products.
Amazon says Meta did not get permission and that Muse did not clearly identify itself as an automated program.
Amazon is also worried that the tool could keep customer information or make wrong or bulk orders.
Meta says Muse cannot directly read users’ passwords or banking information.
Meta says login details are kept in a protected environment.
Muse users are automatically included in a program that can use their conversations to improve AI models.
Users can turn that setting off in Data Controls.
The disagreement reflects a wider argument about how AI agents should access websites and make purchases.
Amazon blocked Meta’s Muse AI agent from browsing or purchasing products on Amazon.com.
Amazon said Muse accessed its website without consent and violated the retailer’s Conditions of Use.
Amazon had reportedly asked Meta to exclude Amazon voluntarily before imposing the block.
Amazon raised concerns about automated identification, customer credentials, retained information, and incorrect or bulk orders.
Meta said Muse cannot directly access passwords or banking data, while users can disable the use of interactions for AI training.
- Who
- Amazon and Meta, concerning Meta’s Muse AI agent.
- What
- Amazon blocked Muse from browsing and purchasing on Amazon.com over alleged unauthorized access and privacy and security concerns.
- When
- The block was reported by Sunday night, September 20; the articles do not specify the year.
- Where
- Amazon’s website and online stores, including Amazon.com.
- Why
- Amazon said Muse had not been authorized, appeared not to identify itself as automated software, and could create risks involving credentials, retained customer information, and incorrect or bulk orders.
Amazon’s Concerns
Meta’s Response
Authorization and transparency
Amazon’s Concerns
Amazon said it never consented to Muse’s integration, that Meta launched it without prior notice, and that third-party applications should transparently respect a business’s decision about whether to participate.
Meta’s Response
Meta has emphasized Muse’s ability to perform tasks for users and has defended its safeguards, but the articles do not report that it had obtained Amazon’s permission.
Credential security
Amazon’s Concerns
Amazon warned that Muse appeared to collect and retain customer credentials, creating serious privacy and security risks.
Meta’s Response
Meta said Muse does not directly access passwords or banking information and that login details entered by users are stored in a protected environment.
Data and automated orders
Amazon’s Concerns
Amazon said retained customer information or errors by an automated agent could lead to incorrect or bulk orders.
Meta’s Response
Meta said interactions used for AI training are scrubbed of identifying information and that users can turn off the training-data setting.
- Blocked service
- Meta’s Muse autonomous AI agent
- Affected platform
- Amazon.com and Amazon’s online stores
- Amazon’s stated rule
- Unauthorized AI-agent access violates Amazon’s Conditions of Use.
- Prior Amazon action
- Amazon reportedly asked Meta to exclude or restrict Amazon voluntarily before blocking Muse.
- Credential safeguards claimed by Meta
- Meta said Muse does not directly access users’ passwords or banking data, and login information is stored in a protected environment.
- AI-training setting
- Users can disable the “Help improve our AI models” setting in Data Controls.
- Broader business ties
- Amazon has allowed purchases through Facebook and Instagram feeds since 2023, and Meta signed a contract in April to run autonomous AI agents on Amazon’s Graviton server chips.
