Meta's Muse Goes Viral but Blocked by Amazon: AI Shopping Agents Spark Platform Control Battle

Meta Platforms (META-US) has launched its personal AI agent Muse, which has quickly gained popularity, ranking high in the Apple App Store's free app charts in the U.S. and driving Meta's stock price up by over 20% in just two weeks. However, as Meta CEO Mark Zuckerberg prepares to showcase the latest AI and wearable devices at the Meta Connect conference, Amazon.com (AMZN-US) has blocked Muse from making purchases on its platform, raising questions about whether AI agents can execute tasks across platforms and who will control the direct relationship between consumers and businesses.

Meta released Muse on September 8, positioning it not as a traditional chatbot but as a personal AI agent that can actually execute tasks on behalf of users. Meta states that Muse can help with scheduling appointments and services, filling out spreadsheets, monitoring home security cameras, and even searching for products and completing purchases. Users only need to state their needs, and Muse will search for products across multiple websites and complete transactions through payment and e-commerce services like Shopify (SHOP-US) or Stripe.

Muse's rapid growth has gained Meta significant attention in the consumer AI market. CNBC reports that Muse jumped to the top of the Apple App Store's free app download chart in the U.S. in less than two weeks, even surpassing ChatGPT. Meta's stock price also rose by over 20% in the two weeks following Muse's launch. As of September 23, Meta's stock price has risen by approximately 13% so far this year, slightly below the Nasdaq index's approximately 16% gain over the same period.

However, one of Muse's most important functions is precisely the area that Amazon is least willing to open up. Amazon has recently blocked Muse from making purchases on its platform on behalf of users, believing that third-party AI applications representing consumers must operate transparently while respecting whether service providers are willing to participate. Amazon also stated that Meta did not notify Amazon in advance that Muse would access its e-commerce website and did not obtain Amazon's authorization.

When users attempt to shop on Amazon through Muse, they will see Amazon's prompt message stating that unauthorized AI agents are continuously accessing Amazon services, violating the terms of service. Amazon's move effectively cuts off Muse's shopping capabilities on its platform.

Amazon's reasons go beyond platform authorization and also include security and account data issues. Amazon pointed out that Muse did not clearly indicate that it was an automated agent while browsing the website, and Amazon believes that the agent program may capture and store user login credentials, thus posing privacy and security risks.

This controversy highlights an increasingly important issue between AI agents and traditional internet platforms: as AI begins to replace consumers in searching, comparing, ordering, and paying, will consumers still directly enter large platforms like Amazon? If shopping entry points gradually shift to AI assistants, e-commerce platforms that have controlled consumer traffic, ad exposure, and transaction processes may need to rethink their role in the shopping process.

eMarketer analyst Max Willens stated that as more consumers start using AI agents like Muse, Amazon may place more emphasis on its role in the entire "consumer journey." He pointed out that over the past few years, Amazon has been an important first stop for U.S. consumers shopping online, but now some consumers may start initiating the shopping process through AI assistants.

This is not the first time Amazon has clashed with AI agents. Amazon has previously taken restrictive measures against other AI services, including legal disputes with Perplexity. Amazon has also blocked agent-type tools from OpenAI and Google. CNBC reported that Amazon has previously shown an attitude of not wanting AI services to form a new intermediary layer between itself and consumers.

Smart alarm clock company Loftie and AI e-commerce startup Deliberate CEO Matthew Hassett stated that one of Amazon's core business models is to firmly grasp the consumer relationship, so Amazon's move against Muse may not be just a simple technical or security issue, but a larger competition over who controls the consumer shopping entry point.

In contrast to Amazon's blockade, Shopify has taken a completely different strategy. Shopify announced a collaboration with Meta, allowing Muse to complete agent-style checkout through Shop Pay in Shopify stores. This means that Shopify did not try to block AI from intervening in the shopping process but chose to become an important payment and business infrastructure when AI agents complete transactions. After Shopify announced the collaboration, its stock price temporarily rose significantly, with a closing increase of over 7% on September 21.

Muse may therefore form two completely different development models. On the one hand, large platforms like Amazon may choose to block third-party AI agents to maintain their direct relationship with consumers; on the other hand, Shopify chooses to integrate with AI agents, hoping to gain transaction and payment opportunities from the new agent shopping process.

The impact of AI agents is beginning to extend from e-commerce to other industries. CNBC noted that the task execution capabilities of Muse have attracted the attention of financial services and travel industry players. Recently, the stock prices of Charles Schwab (SCHW-US) and LPL Financial Holdings (LPLA-US) have fallen, while the stock prices of Booking Holdings (BKNG-US) and Expedia Group (EXPE-US) have also attracted market attention, reflecting that investors have begun to assess whether AI agents may directly intervene in the service entry points originally controlled by these platforms.

For Meta, Muse is not just a new app. Meta has primarily controlled user attention through social platforms like Facebook and Instagram, with revenue highly dependent on digital advertising. If Muse can become the primary entry point for consumers to handle daily tasks, Meta has the opportunity to evolve from "controlling attention" to "controlling task and transaction entry points." This is also an important reason why the market has begun to re-evaluate the commercial value of Muse.

Cantor analyst pointed out that Muse's current unit economics may still be affected by early subsidies, but Meta has multiple ways to build a profit model, including freemium models. However, Muse's real commercial model is not yet fully established, and Meta's existing revenue is almost entirely from digital advertising, so how to convert the massive AI usage into sustainable income remains a focus for future observation.

Meta Connect will become an important stage for the next phase of Muse's development. Zuckerberg is expected to take the stage at the annual developer conference, and one of the market's focuses is whether Meta will further integrate Muse with wearable devices. Mizuho analyst believes that Meta Connect may shift the market's understanding of Muse from simple app adoption rate to the "platform" concept, especially the possibility of integrating AI agent capabilities into Meta's smart glasses.

Meta has not formed a consumer market of the same scale as smartphones with its Quest VR products in the past, but its smart glasses co-developed with EssilorLuxottica, the parent company of Ray-Ban, have recently attracted relatively more attention. If Muse is integrated with smart glasses, users may be able to directly request AI to complete tasks such as searching, shopping, booking, and others through wearable devices in the future, and the usage scenarios of AI agents will further extend from smartphone apps to daily life.

However, whether Muse can develop from the current download craze into a sustainable platform still depends on multiple factors, including users' trust in privacy and security, whether retailers are willing to open up platforms, the attribution of responsibility when AI agents execute transactions, and whether Meta can build a sustainable business model. Amazon's blockade of Muse also shows that when AI begins to attempt to replace part of the search, recommendation, and transaction entry points of platforms, the competition between large internet companies may extend from models and computing power to consumer relationships and transaction control rights.

For the entire AI industry, Amazon and Meta's stalemate may just be an early case of the agent business era. In the future, when AI agents from Apple, Google, OpenAI, and other tech companies also have the ability to search, order, and pay across multiple websites, whether major platforms choose to open up, negotiate cooperation, or block third-party agents may all affect how AI reallocates network flow.