Meta Platforms introduced its new personal artificial intelligence agent called Muse on September 8, and it has quickly gained widespread attention in the tech world. The app climbed to the top of Apple’s U.S. App Store and passed competing AI apps including ChatGPT.
Muse had about 730,000 downloads by September 18, according to Sensor Tower data cited by Yahoo. That was about 10 days after its release rather than five days. More recent figures show its growth continued rapidly. Reuters reported on September 22 that the app had been downloaded more than 2.5 million times. Another Reuters report cited 2.8 million downloads during its first 12 days.
The strong early performance of Muse has contributed to a significant rise in Meta’s stock value, which climbed over 11% on Monday. Reuters reported that Meta gained roughly $200 billion in market value as enthusiasm around Muse grew. Meta shares had risen more than 20% since Muse launched by September 22.
Muse is designed not only to answer questions but also to carry out tasks such as filling out forms, booking travel, sending emails, and shopping on behalf of users. Meta describes it as a personal AI agent rather than a traditional chatbot. It can browse the web, make purchases, create documents and connect with outside apps and services.
The app offers both a free version and paid plans for heavier users, with monthly fees of $20 or $100 depending on the service level. Muse launched in the United States through its own app and through WhatsApp.
Meta’s launch of Muse marks a major push into consumer AI agents. The company is competing in an increasingly crowded field that includes OpenAI, Anthropic, Google and other companies building software capable of taking actions for users.
Despite its popularity, not all tech companies are welcoming to Muse’s approach. Amazon blocked Muse from using its e-commerce platform to make purchases for customers. Users attempting to shop on Amazon through Muse began receiving messages saying that access by an unauthorized AI agent violated Amazon’s Conditions of Use.
Amazon said Meta had not informed the company before Muse began accessing its shopping site. Amazon also said Muse did not clearly identify itself as an automated agent and raised concerns about whether customer credentials could be captured or stored. Meta has said Muse cannot access secure login or payment information in the way Amazon suggested.
The dispute also has broader business implications. AI agents could change how people shop by allowing software to search products and make purchases without users moving through traditional websites. That could affect advertising, product placement and the direct relationships companies have with customers.
Meta has partnered with Shopify to support shopping through Muse. Shopify is integrating its Shop Pay checkout system with the agent. Shopify shares rose after the partnership was announced as investors reacted positively to the company’s role in agent-based shopping.
Amazon is taking a different approach. The company has its own AI-powered shopping tools. Rufus, Amazon’s conversational shopping assistant, was renamed Alexa for Shopping in May 2026. It can recommend products, compare prices and perform some automated shopping tasks.
Amazon also operates a feature called Buy for Me. The service can purchase some products from outside brand websites when those products are not sold directly through Amazon. Brands have a choice over whether to participate.
Meta’s Muse app is part of a broader trend toward AI agents that go beyond simple search engines to perform real-world actions on behalf of users. That greater ability has also brought new questions about security and privacy.
One security concern became public shortly after the launch. Security researcher Patrick Wardle discovered a vulnerability in the macOS version of Muse that could allow an attacker who already had malicious software running on a user’s computer to redirect Muse and potentially cause it to perform unwanted actions. Meta patched the vulnerability after it was reported. The company said the real-world risk was limited because an attacker would first need local access to the computer.
Meta has also been testing a feature in which human contractors can assist Muse with telephone calls. Reuters reported that some Meta employees raised privacy concerns because those calls could expose sensitive information to the human workers. Meta said the testing was intended to help the company improve safety and privacy before making the feature broadly available.
Meta says Muse includes safeguards designed to protect users. The company says its system uses a dedicated virtual machine for web activity and keeps sensitive information encrypted. Users can also control which apps and services they allow Muse to access.
Some consumers are still hesitant to let AI tools handle purchases, messages and other personal tasks directly due to privacy and trust issues. The more actions an AI agent can perform, the more important permissions and account security become.
Despite these concerns, Muse’s rapid rise in popularity suggests a growing interest in AI tools that can manage everyday tasks efficiently. Wall Street analysts are already examining whether Muse could become a major new source of revenue for Meta. Jefferies estimated that the service could generate $10.8 billion in annual revenue by 2027 under one scenario involving one billion users and a small share paying for subscriptions. That estimate is an analyst projection rather than company guidance.
Muse’s early success has also strengthened Meta’s position in its competition with companies such as OpenAI and Anthropic. Meta already reaches billions of people through Facebook, Instagram and WhatsApp, giving the company a large existing audience for promoting new AI products.
As the market for AI agents continues to evolve, questions remain about how user data will be protected and who controls access to online services. The disagreement between Amazon and Meta shows that websites may not automatically allow outside AI agents to browse and transact on behalf of customers.
Meta’s strategy with Muse may influence future developments in consumer AI tools and the way companies interact with each other in this space. Its early popularity shows strong demand. The Amazon dispute and recently patched security flaw also show some of the challenges that come with giving AI software the ability to take actions on a user’s behalf.
IMAGE: AMRULQAYS Pixabay