Amazon has begun blocking the Muse AI agent from Facebook parent Meta, after Meta declined a request to voluntarily remove Amazon from the bot’s activities.

The company said it began blocking Muse on Sunday. The site now displays a warning message saying use by an “unauthorized AI agent” violates its terms of use with customers.

Amazon said it believes agents should “operate openly” and should allow companies to opt out of participation, adding that Muse’s apparent use of users’ Amazon credentials could pose a security risk.

Popular app

Meta has said previously that credentials are stored in such a way that the agent can use them without directly accessing them.

If users shift to other platforms to buy from Amazon, the company stands to lose advertising revenue, but barring bots from its service could also mean a loss of sales.

The Muse app has quickly become popular with consumers, rising to the top of the Apple and Google app store charts for free apps as of Monday.

Meta and Amazon work together in some AI and e-commerce ventures, with Facebook and Instagram users able to purchase Amazon products from inside their apps, and Meta using Amazon’s Graviton chips to run agent workloads.

Amazon is currently in the midst of legal action with Perplexity AI to bar Perplexity’s AI agents from its e-commerce site.

Shopping bots

Amazon initially claimed Perplexity’s technology was effectively hacking its service, but following a 4 August ruling that went in favour of Perplexity, it has shifted its claims to focus on terms of service agreed by the user.

Amazon’s own automated shopping tool, Buy For Me, has been criticised for listing products from independent vendors, in many cases without their knowledge, forcing vendors to contact the company to be removed from the service.

The Seattle-based tech giant says it allows vendors to opt out of Buy For Me.