San Francisco: (RightNow) Meta’s AI agent ‘Muse’ has been revealed to use human contractors to complete phone calls. Some calls made through Muse were actually conducted by humans, raising privacy concerns for users.

According to internal posts from Meta, the ‘Human Concierge’ experiment aimed to enhance AI’s phone calling success. However, it raised concerns about user data protection. Muse is designed to perform daily tasks based on user instructions.

Some businesses were not accepting AI calls, leading Meta to employ human assistants. Employees at certain organizations hung up on Muse’s calls, making task completion difficult.

It is clear that Meta had emphasized user security, but the involvement of human workers has cast doubt on those promises. The company has currently suspended this experiment, but efforts to improve AI’s calling capabilities will continue.

It is noteworthy that Meta previously tested a digital assistant named ‘M’, where humans often performed tasks. The current experiment highlights the need for privacy and transparency to make AI effective in practical life tasks.