Meta’s personal AI agent ‘Muse’ has quickly become a major talking point in the AI world since its launch earlier this month.In just the last 24 hours or so, Muse has climbed to the top of mobile app charts, rallied Meta’s stock, gotten blocked by Amazon for allegedly violating its terms of service, and drawn scrutiny from researchers who say they discovered a zero-day vulnerability that could potentially allow hackers to hijack the AI agent.Powered by Meta’s large language model (LLM), Muse Spark, the personal AI agent app has been pitched as a way for consumers to delegate a number of tasks across the web, like filling out electronic forms, organising email inboxes, booking travel tickets, turning recipes into grocery lists, sending party invitations, and making purchases at checkout.Muse is able to execute these tasks by using its built-in connectors to navigate through a user’s email inbox, calendar, payments apps, health and fitness apps, smart home device apps, e-commerce apps, dining apps, and more.The macOS app (there is no Windows version yet) also works with a user’s WhatsApp and social media accounts. When a task requires a tool that does not exist or a locally installed app does not have a supporting connector, Muse creates one on the go. While the Muse app is free to download, Meta is offering in-app monthly subscriptions that cost $20 or $100, depending on usage.Muse did not generate a lot of buzz at launch. However, as use cases emerge and people share their experiences online, Muse seems to be gaining ground in the increasingly crowded market for autonomous AI agents. While OpenClaw similarly exploded in popularity among developers and tech enthusiasts earlier this year, Muse is positioned more as a personalised, consumer-facing AI agent.It is Meta CEO Mark Zuckerberg’s latest bet on AI to generate revenue beyond the social media giant’s core advertising business and perhaps start recouping some of the billions of dollars invested in AI infrastructure and talent.Story continues below this adBut, why is Meta’s Muse agent suddenly making headlines? Why has Amazon blocked it from visiting its shopping site? What does this mean for the future of agentic commerce? How much access does Muse have to a user’s device and accounts? And could a security flaw potentially let hackers take control? Take a look.The rise of Muse, in numbersAs of Monday, Muse is ranked as the No 1 free app on the US version of Apple’s App Store and Google Play Store. In the six days following its release on September 8, the app was downloaded more than 9,02,000 times, according to data from Sensor Tower. The Meta AI app, in comparison, only saw 1,29,000 downloads in the same post-launch period.Muse is also ahead of ChatGPT as the leading free-to-download iOS app in the US. It is also ahead of rival AI chatbot apps such as Anthropic’s Claude and SpaceXAI’s Grok AI along with other popular services like Polymarket and Kalshi. The market intelligence firm further estimated that Muse has seen more than 2.6 million downloads since its debut, with 1.5 million and 1.1 million downloads on the iOS and Android app stores, respectively.Over the same time period, Claude and Grok chatbot apps only recorded 4,00,000 and 2,00,000 downloads, respectively. Muse’s cumulative downloads is second only to ChatGPT (3.1 million downloads in 13 days post-launch), as per Sensor Tower.Story continues below this adAlso Read | Meta launches Muse AI agent: What it can do, and how it handles your dataNotably, the other apps are primarily AI chatbots, while Muse is the only one in these ranks that offers access to a personalised AI agent. This signals growing interest in AI agents beyond the developer community. Most agentic AI use cases so far have not been aimed at normal users, but it appears that Meta’s Muse and similar offerings are seeing more broad-based adoption.The potential of growing consumer demand for Muse was also reflected in Meta’s stock, which surged more than 11 per cent to $741, its highest closing price in a year. An analyst cited by The Wall Street Journal estimated that Muse could add $28.5 billion to the company by 2030, presumably in revenue. Meta’s revenue last year was $200 billion.Amazon draws the lineThe growing consumer interest and revenue potential of agents like Muse depends on how much access it can get to the websites and services that people use online. This has emerged as an early point of friction.E-commerce giant Amazon blocked Muse from accessing its shopping site, citing privacy and security risks. Users who tried to direct the AI agent to Amazon received a message stating that Muse was an “unauthorized AI agent [that] violates Amazon’s Conditions of Use.”Story continues below this adDefending its move, Amazon said it did not receive advance notice from Meta that its AI agent would be accessing its online store. The company alleged that Muse is able to bypass the personalisation and other features built into the Amazon shopping experience, and could capture and store customer credentials and scrape account data.“We think it’s fairly straightforward that third-party applications that offer to make purchases on behalf of customers from other businesses should operate openly and respect service provider decisions about whether or not to participate,” an Amazon spokesperson said in a statement.“This helps ensure a safe, secure, and reliable customer experience, and it is how others operate including food delivery apps and the restaurants they take orders for, delivery services apps and the stores they shop from, and online travel agencies and the airlines they book tickets with for customers. Agentic third-party applications such as Muse have the same obligations, and we’ve requested that Meta remove Amazon from the experience,” they added.Amazon’s block on Muse underscores a bigger battle brewing among online service providers and consumer AI agent vendors. In the past, Amazon has taken similar action against Google and OpenAI’s shopping bots. Its blocking of Perplexity’s agent led to a lawsuit by the AI search startup, with proceedings still underway.Story continues below this adAlso Read | Why Amazon is trying to block Perplexity’s AI agent from accessing its online storeHowever, if personal AI agents are more widely adopted, consumers themselves could start pressurising online services, including restaurants, grocery delivery apps, content companies, and others, to enable greater access to these agents, according to Palo Alto Networks CEO Nikesh Arora. In a post on X, Arora alluded to a future where people do not visit individual apps or websites anymore, instead relying on their agents to do everything.He also suggested that Amazon is in the same position as news publishers and enterprise software firms in fighting AI firms over control of its site and the underlying data. This standoff, however, may not last long as companies could eventually strike deals with AI firms to accommodate their agents.For instance, Shopify announced Monday that it is partnering with Muse to allow agentic checkout in its online stores. Zuckerberg believes that Muse can make money by “taking a small cut of transactions,” according to Alex Heath’s Sources newsletter. In previous earnings calls, Amazon CEO Andy Jassy has indicated that the company is in talks with other firms wanting to run agents for commerce.