On September 23, according to internal company posts seen by Reuters, Meta Platforms is testing a 'human concierge' mode for its new AI personal assistant Muse, where some calls made by the digital assistant are discreetly answered by real human agents. Last week, Meta introduced this test to its employees, following the public launch of Muse's calling feature. Some employees have raised concerns that having human agents handle calls might lead to the leakage of sensitive information to call center contractors. In response, a Meta spokesperson stated that employee feedback has been 'overwhelmingly positive,' and the test aims to gather user opinions, refine safety and privacy measures, and improve functionality. The spokesperson also mentioned that Meta is collaborating with businesses to enhance this calling feature and will only launch it once the feature is ready and accompanied by appropriate disclosures.