Analysis
Meta's AI companion app Muse has topped 3 million weekly users, according to an exclusive from The Information, even as the company disputes a user's claim that Muse read their private messages without permission, per TechCrunch.
What's Changed Since We Last Covered Muse
Pulse previously covered Meta's push to open-source parts of Muse's underlying model. The new development is twofold: a first hard usage number -- The Information's first disclosure of real engagement data for the product -- and a fresh privacy controversy that lands at an awkward moment for a consumer AI companion app trying to prove it can be trusted with personal conversations.
The Adoption Problem
Axios reported that AI agents broadly -- Muse included -- still have what it called a "normal-people problem": heavy use among early adopters and AI enthusiasts hasn't yet translated into mainstream habitual use the way ChatGPT has. Three million weekly users is a real number for Meta to point to, but it's a fraction of the daily active base Meta's other apps -- Instagram, WhatsApp, Facebook -- each command in the billions.
The Counterweight
A privacy dispute over whether Muse read private messages without permission -- however it's ultimately resolved -- is exactly the kind of story that slows mainstream adoption of AI companion apps rather than accelerating it. Meta's denial doesn't settle the question technically; it's a public dispute between the company and a user, and the underlying facts of what the model accessed and why remain unresolved in current reporting.
What to watch: whether Meta discloses daily-active or retention numbers to back up the usage milestone, and how the privacy dispute resolves -- the two threads are now tied together in how outsiders judge whether Muse deserves the access to personal data it already has.
