Meta’s new personal assistant, Muse, has become a viral hit, with millions downloading the AI agent, connecting it to bank accounts, messages, or health data, and allowing it to complete tasks for them. But as Muse takes off among consumers, data from inside the app is providing insight into how it organizes and presents information to users.
\n\nFor more like this subscribe to “Kernel Panic,” a new weekly newsletter by Lily Hay Newman and Matt Burgess from inside the world of privacy and digital security.
\n\n[Matt] Millions of people have downloaded Muse,
\n\nMeta's AI agent with a cute and cuddly looking mascot.
\n\nthat getting Muse to complete task for you
\n\nAn independent AI safety and security researcher
\n\nextracted Muse's operating instructions and safety prompts
\n\nto copy and share its own system files.
\n\nwhich the researcher shared with Wired,
\n\nappears to be the ability to create a page
\n\nfor every person in the user's life.
\n\nWhat this means is that Muse compiles data
\n\non a user's family and friends, even the people they follow.
\n\nIt can then make suggestions like giving pointers
\n\non how to improve particular relationships.
\n\nThe security researcher likened this
\n\nto Muse wanting to know the user like…
Original source: https://www.wired.com/