Meta Tests Muse AI Agent Calls That Are Made by Humans in a Call Center
10 hours ago
- Meta's AI assistant Muse, advertised as able to call businesses, is secretly using human call-center agents for some calls, according to internal communications seen by 404 Media.
- Internal Meta posts confirm a 'human agent layer' was added to complete calls, with human agents handling requests when needed, despite public claims of AI autonomy.
- Employees raised concerns about privacy (sensitive data shared with human contractors) and negative PR, warning that news of human involvement could damage trust and be framed as 'AI isn't good enough'.
- Some testers were not informed right away that a human made the call on their behalf, heightening worries about user expectations and consent.
- Meta defended the practice as internal testing, saying it is working on safety and privacy protections, and will only launch publicly with proper disclosures.
- This is part of a broader pattern where companies claim AI automation but rely on human labor behind the scenes, risking credibility and public backlash.