Muse's 'call on your behalf' feature exposed as partly relying on real humans; Meta has rolled back the test.

Beating AI News Flash: Just as Meta expanded testing for Muse's "call on your behalf" feature, it was revealed that some calls were actually completed by humans.

Internal information obtained by 404 Media and Reuters shows that Muse can transfer tasks to human contractors at call centers, who make calls and complete requests on behalf of users. Some testers only learned after the call ended that a human had intervened.

Meta mainly did this to improve the success rate. Internal company testing shows that with human intervention, the success rate for phone tasks can reach 95% to 98%, higher than pure AI.

But one practical problem is that some businesses hang up immediately upon hearing that the other party is an AI. Some testers in the community also said that businesses hung up after Muse identified itself as an AI assistant.

Human intervention also raises privacy concerns. Users originally thought their information was only being given to AI, but in reality it may be seen by outsourced personnel. Some people inside Meta also questioned that giving contractors privacy training does not amount to a real security mechanism.

A Meta executive later admitted that failing to clearly tell testers in advance that humans would intervene was a mistake, and the feature has been rolled back. Meta said the phone feature is still in testing and will only be officially launched in the future after security, privacy, and clear disclosure are properly addressed.