Muse hands its calls to a trained human agent.
Callers cannot tell whether an AI or a person is actually speaking to them.
Meta pitched Muse as autonomous, but a human agent places the call and files a transcript afterward.
An internal Meta message-board post warned the feature “could portray us as their AI is not good enough.”
Meta has not disclosed how often Muse calls are human, and employee privacy concerns remain unresolved.
How each outlet framed it
- 404 Media leans critical
- reveals Meta's public AI-call claims mask internal human call-center layer via internal company posts
Sources: 404 Media