On October 6, 2026, Citigroup published a research note that put hard numbers on Meta's biggest AI bet yet. The brokerage estimates that Muse — Meta's personal AI agent assistant — could generate more than $27 billion in annual revenue by 2030, split between roughly $23 billion from transaction-related revenue and about $4.5 billion from subscriptions.
Muse was unveiled last month and launched on September 8, 2026. Citi says it has already topped 6.6 million cumulative downloads, with 1.8 million daily active users as Meta ramps up marketing and new use cases emerge. Those are early numbers, but the trajectory is why Wall Street is paying attention: Meta's stock is up more than 20% since Muse debuted, pushing the company within striking distance of a $2 trillion market value.
What "front door to the internet" actually means
Citi's phrase is doing a lot of work: personal agents, it argues, are emerging as the new front door to the internet. The idea is simple. Today, you open a browser, type a search, and click through websites. Tomorrow, you might just tell your agent what you want — and it does the searching, comparing, and buying for you.
That shift is the whole bet. If Muse becomes the thing billions of people talk to before anything else, Meta owns the layer between users and the entire internet. Whoever controls that layer can take a cut of commerce, sell subscriptions, and keep users inside their ecosystem instead of Google's or Apple's. It is a first-mover race, and Meta is running against OpenAI's ChatGPT Dots, Google's Gemini, and Anthropic's Claude — all building their own agents.
How Meta plans to win the living room
Meta is not just building software. According to recent reports, the company is preparing to ship Muse Home Link in the coming weeks: a small USB-powered dongle that connects Muse to a home network so it can talk to TVs, speakers, and any smart-home device that exposes a web connection. The company is producing a limited initial batch of about 5,000 units — a toe in the water, not a mass launch.
The strategy deliberately echoes the early Android playbook. Meta has open-sourced the Home Link SDK, which already has thousands of stars and hundreds of forks on GitHub and supports hardware from ESP32 boards to Raspberry Pi and Linux systems. If outside manufacturers build Muse-compatible devices, Meta gets an ecosystem without manufacturing every gadget itself. The one caveat: the official dongle runs locked firmware, so Meta is keeping tight control over the core experience even while inviting developers in.
Meanwhile, the pricing runs from a free tier all the way to a $100-a-month power-user plan — a signal that Meta sees agents as a serious productivity tool, not a toy.
Why this bet is so big — and so risky
The logic is seductive. Meta already has billions of users on its apps. If even a fraction of them start their day with Muse, the transaction revenue alone could dwarf the ads business model that built the company. Citi's $27 billion estimate assumes Muse becomes genuinely woven into daily life — shopping, booking, planning, working.
But nothing about this is guaranteed. The competition is fierce and well-funded, and users are famously fickle about new platforms. Apple's own Siri AI agent has reportedly slipped to late 2026 or early 2027, which gives Meta a window — but windows close.
Since the Citi note, Meta has kept pushing: the Muse Wonder model series launched on October 12, aiming to power the Muse app for paying subscribers. The message is consistent — Meta wants its AI to be the most capable agent you can actually buy.
If personal agents really do become the new front door to the internet, the stakes could hardly be higher. And Meta is betting the house that it owns the key.
Would you let an AI agent do your shopping and browsing for you, or do you still want to click through yourself? Tell us in the comments.
Comments
Post a Comment