“Senator, we run ads” -Zuck and eight years later those ads have a butler’s voice and your bank login.
3 min readIn April 2018, Senator Orrin Hatch asked Mark Zuckerberg how Facebook could stay in business without charging its users. The answer became a meme: “Senator, we run ads.” Eight years later, Meta has launched Muse, a personal AI agent that talks to you, remembers your goals, and can open a browser, fill out forms, and pay for things on your behalf. The business model has not changed so much as grown hands.
What Meta Muse actually is
Muse launched on September 8 on iOS, Android, WhatsApp, and muse.ai, and is rolling out in the US first. Meta calls it “a secure, private personal AI agent” that “doesn’t just answer questions, it actually does the work.” Give it a goal and it builds a plan, coordinates your calendar, and keeps working after you close the app. It can send email, book travel, negotiate a bill down, and turn a recipe reel you saved on Instagram into a grocery list.
It runs on Muse Spark, which Meta describes as its most capable model to date, built for agentic work. Pricing is free for most use, with $20 Power and $100 Maximum tiers, according to TechCrunch.
The butler’s voice and the wallet
At Connect on September 23, Meta added the parts that make the headline land. Muse gets a real-time video avatar you can talk to, a wake word on Meta’s smart glasses, its own email address, and the ability to drive any Mac application while you step away. On the money side, Muse is the first AI agent covered by Stripe Link’s purchase protections and pays with a one-time card so your real number stays hidden. Shop Pay, PayPal, and 1Password support for your existing logins are on the way, alongside retail connectors for Walmart, Best Buy, Instacart, Expedia, and others.
Zuckerberg was direct about how this gets paid for: “We’re standing behind this by making Muse free for a huge number of tokens, with the expectation that over time we will profit by taking a small fee from transactions.”
Why it matters
Meta insists Muse has no visibility into your passwords or payment methods, runs inside a dedicated Secure VM, and does not share conversations with its ad systems. A separate Sentinel agent has to approve anything Muse sends to the internet, and a Confidential VM encrypted with a key only you hold is promised later this year. Those are strong claims from a company carrying a 2011 FTC consent decree, a 2019 privacy settlement, and fresh multibillion dollar settlements over harms to children. TechCrunch’s verdict is that the security story “will require deeper investigation by security experts.”
The bigger shift is the incentive. An ad model wants your attention. A transaction fee model wants your agent to buy things. That is a different kind of pressure on a product that already holds your calendar, your inbox, and your card. Watch how quickly “suggests ideas unprompted” starts to look like the old feed with better manners.