Meta's new consumer AI agent Muse took the spotlight at this year's Connect event, where Mark Zuckerberg made clear the company intends to push AI into everything it ships — up to and including a Tamagotchi-style AI device that Meta insists is for adults only. On the latest TechCrunch Equity podcast, reporters asked the obvious follow-up: will anyone trust Meta with the data Muse needs?
Sean O'Kane, who has been using Muse since launch, had one genuinely delightful moment. The agent suggested scanning for unclaimed property funds in his name; to his surprise, there were some, and a check is now in the mail. "It helped me make some money on my first day," he said. But he was quick to add a caveat: that is a one-time event, "a party-trick type thing", not a repeatable daily habit.
The durable version of the pitch is what Meta has been promoting lately — connecting Muse to your real financial life: a credit card, your Gmail, your subscriptions, so it can cancel what you no longer use and flag double charges. That is exactly where the trust wall appears. "Meta's business is to sell you ads," O'Kane said. He added that he would hand that same access to Apple's Siri far more readily, because he trusts that company both on security and on incentives.
One detail cuts the other way. On first install, Muse did not immediately pull in his Instagram, Threads and Facebook context, which made it feel less invasive than expected — although it gradually pushes to absorb more data. Meta says it deliberately delayed Muse's release to improve safety, security and privacy; other reporting has raised questions about the agent's deep permissions and its reach into the file system.
Strategically, Meta is swimming against the current. OpenAI and Anthropic have spent the week on coding and enterprise tools, chasing the revenue those markets promise. Meta is betting that consumer distribution — more than a billion existing users — is the moat that matters, and that a friendly, everyday assistant can win where an enterprise SKU cannot.
