Meta's Muse AI agent surges to top app charts amid security fixes

Meta's new Muse AI agent reached number one on U.S. app stores within days of its September 8 launch, outpacing early downloads of ChatGPT and TikTok and contributing to an 11 percent rise in Meta shares. The tool, positioned as an autonomous assistant that can review finances, cancel subscriptions, negotiate bills, and handle shopping tasks, drew millions of users who shared savings examples under a company-promoted hashtag. Meta chief AI officer Alexandr Wang highlighted cases such as one user recovering a year-old book subscription refund and another cutting annual car insurance costs by $3,500. At the same time, Amazon blocked Muse from completing purchases on its platform, citing violations of terms of service. Security researchers identified a zero-day flaw in the macOS version that allowed local attackers to redirect cloud transcription and issue commands through the agent; Meta released a patch within hours and stated the issue required prior malicious code on the device. Broader AI hardware stocks rose alongside Meta's gains, though independent verification of typical user savings remains limited to individual reports. The rapid rollout has also surfaced questions about how such agents interact with existing platforms and whether design choices prioritizing cloud processing increase exposure.