A personal account published by multiple outlets says the writer gives day-to-day tasks to Meta’s AI agent “Muse,” reporting that it successfully handles items such as managing dental insurance, booking dinner reservations, and creating a podcast.

The articles also emphasize the trade-off they see in doing so: the agent requires access to highly personal data, and the writer frames their experience as a test of trust. While all sources focus on the same general storyline—how the AI agent carries out practical errands and creative work—their presentation centers on the user’s perspective rather than independent testing or broader results. Some coverage highlights the convenience of delegated tasks, while other attention is directed to concerns about privacy and the amount of information used. Overall, the outlets present the experience as both impressive in capability and sensitive in terms of data access and user confidence, without detailing specific technical specifications or verified performance metrics.