Meta patches Muse flaw that could expose AI agent accounts
Meta has released a patch for its Muse macOS app after security researcher Patrick Wardle identified a zero-day vulnerability. The flaw required an attacker to have local access to the user’s device, but it could potentially allow control of the Muse AI agent and access to the associated Muse account.
Wardle found that an undocumented Muse setting allowed code running locally to redirect transcription processing away from Meta’s servers to an endpoint controlled by an attacker. Ars Technica reported that this could give the attacker access to the account. The vulnerability was linked to several design choices: Muse handled dictation in the cloud rather than on the device, and any application could control all of the app’s undocumented settings. Meta’s patch addresses the issue in the macOS application.
Written by Sakhanda Wire AI from reporting by The Verge. Facts, quotes and full context are in the original article below.