Meta Centers Muse on Rapid Adoption, Deeper Access and Privacy

At Meta’s 2026 developer gathering, Mark Zuckerberg said the company’s future lay in Muse, a personal agent that was not yet a month old. [1] The emphasis follows earlier Meta platform bets: Facebook released a Messenger bot platform in 2016, began winding it down by 2018, and later pursued a cryptocurrency project that never came to fruition. [1] Meta is now promoting Muse at its annual Connect conference after early reviews praised the application’s polished design and the capabilities of its underlying AI model. [2]

Sensor Tower estimated that Muse rose from more than 2.5 million downloads at the start of the week to more than 3.4 million by Thursday, following its September 8 launch. [2] Competing estimates varied substantially: Apptopia counted 4.3 million downloads, including 2.6 million on iOS, while Appfigures reported roughly 2.3 million installs divided almost evenly between iOS and Android. [2] Muse was available only in the United States and Canada, and Sensor Tower measured a 27 percent increase in daily active users on the Wednesday after Meta Connect concluded. [2]

Meta announced planned Muse features including video conversations with its avatar, Mac computer use, a dedicated email address, additional partners and connectors, and smart-glasses integrations. [2] The Verge also described Meta as making Muse’s filesystem more accessible. [3]

Meta is seeking to make privacy protections central to Muse and other new products because personal assistants may receive emails, financial records, health data and other sensitive information. [4] Muse gives each customer a personal virtual machine, and Meta says a forthcoming option would prevent even the company from accessing its data. [4] Meta is also preparing a similar private-processing option for interactions with its AI glasses, following earlier incognito-chat and private-processing work across WhatsApp and other Meta applications. [4]

Muse combines three elements Meta’s earlier Messenger bot initiative lacked in lasting form: rapid consumer uptake, a broad set of planned actions and a privacy architecture intended for sensitive data. The comparison that matters is ChatGPT, because Axios describes Muse as potentially becoming one of the most popular AI hits since that product. Yet the conflicting download estimates show why adoption should be assessed through sustained activity as well as installs, while computer use and filesystem access raise the importance of Meta’s promised private mode. What the sources don't address: when the private option will arrive, what it will cost, and whether independent reviewers will be able to verify its protections.