Meta's AI assistant Muse faces privacy backlash, with users concerned about data access and past scandals.
Meta's AI assistant raises privacy concerns
Meta's new AI assistant, Muse, has drawn criticism over its extensive access to users' personal data, including messages, calendars, and bank accounts. Critics argue this access could enable psychological profiling for targeted advertising, despite Meta's claims of privacy protection.Muse's functionality and user base
Muse promises to assist with daily tasks such as making calls, booking hotels, and planning events. Since its launch, it has been downloaded at least five million times and claims to have three million weekly users. Meta has been actively promoting the assistant, especially in the San Francisco Bay area.Zuckerberg's vision for AI
Meta's CEO, Mark Zuckerberg, envisions a future where superintelligent AI is accessible to all, as detailed in a 6,500-word essay he published recently. This vision has drawn criticism from journalists and authors who question the implications of handing over personal cognitive tasks to AI.Past data misuse by Meta
Meta has faced accusations of facilitating data collection in the past, notably during the 2016 U.S. election, when the company allowed Cambridge Analytica to access data from 87 million Facebook users. Meta later paid $725 million to settle a class-action lawsuit related to this data misuse.Security concerns with Muse
Security experts have raised concerns about Muse collecting detailed user profiles from photos and messages. Researcher Karan Joshi called this approach "particularly worrying," while Meta's spokesperson defended the need for such access to ensure the assistant's effectiveness.Technical flaws and unauthorized access
A technical columnist discovered that Muse for Mac accessed private messages without proper permissions and suggested article ideas based on private conversations. Meta denied these allegations, but a security researcher found a zero-day vulnerability that could allow attackers to bypass permissions and access sensitive data. This flaw was fixed in late September 2026, but doubts about Muse's privacy protections remain.