What happened
Internal company posts show that Meta Platforms is testing a 'human concierge' model for its new artificial intelligence personal assistant, Muse.
Under this model, human customer service staff quietly answer some calls placed through the digital assistant.
Why it matters
The arrangement suggests that some interactions presented as handled by an AI assistant may actually involve human agents, which could raise questions about transparency if users are not informed.
It also indicates that Meta is still working out how to blend automated and human support for its AI assistant.
Key facts
The information comes from internal company posts.
Meta Platforms is testing a 'human concierge' model for its new AI personal assistant, Muse.
Real human customer service staff quietly answer some calls made through the digital assistant.
What to watch next
Whether Meta publicly confirms or explains the human concierge test, and whether users will be told when a human is involved.
How the testing of this model develops for Muse and whether it expands beyond the current trial.
