Meta’s new AI agent Muse has quickly gained attention, presenting itself with a friendly mascot and a reputation for being useful. However, beneath this approachable facade lies a complex surveillance system operated by a powerful tech company known for its lack of accountability.
Muse has been generating buzz, with users praising its capabilities in various aspects of daily life. Eli Tan, writing for the New York Times, referred to his experience with Muse as “the most useful AI app I had ever used,” reflecting the excitement surrounding its launch.
As researchers start to examine this new tool, serious privacy concerns emerge. The most troubling finding is that Muse doesn’t just monitor its users; it also compiles detailed profiles on friends, family, and acquaintances. This means that when users interact with Muse, they unintentionally compromise not only their own privacy rights but also those of the people in their lives.
A report by Wired reveals that Muse is designed to maintain detailed files on individuals connected to a user, frequently updating this information–sometimes as often as once an hour. This constant monitoring raises significant ethical questions about the extent of its data-gathering practices. Karan Joshi, an AI safety researcher who shared insights about Muse with Wired, stated, “What it seemed like to me – from all these prompts, system skills data, and things that they’re feeding into Muse – is that they want to understand your relationships that you have with real people.” He summarized the implications of this intent, saying, “they’re trying to know you like a friend, which is honestly pretty creepy.”
This extensive data collection carries risks. Recently, Meta’s engineers discovered a significant vulnerability that could have allowed users to access the database supporting Muse’s functions. A rushed fix was put in place just before the launch, highlighting the urgency with which the company addressed the issue. An anonymous Meta insider disclosed that “many senior engineers believe it’s inevitable we’re going to have a massive data breach as a result of Muse.”
While it remains uncertain whether the internal files on friends and family would have been part of the compromised database, the risks associated with Muse are clearly substantial. Users must consider not only their own privacy but also that of those around them. This situation raises a pressing question in the age of AI–how far should these technologies go in mapping our social connections?
