Meta, the parent company of Facebook and Instagram, has temporarily stopped a pilot that let human contractors place phone calls on behalf of its new personal AI assistant, Muse. The test, which began last month, was designed to improve the success rate of the agent’s calling feature, which allows users to request tasks such as booking haircuts or checking product availability.
Employees raised privacy worries after learning that human callers could inadvertently receive sensitive user data. One staff member reported that a contractor made a racist remark during a Muse‑initiated call to an internet service provider. In response, a vice president in Meta’s Superintelligence Labs unit said the company had “rolled back this feature” and apologized to the employee.
Meta had highlighted Muse’s safety measures at launch, noting that each agent runs in its own secure virtual machine and stores passwords separately. Critics argued that human involvement could undermine these safeguards. The company’s spokesperson said the test was “overwhelmingly positive” and that feedback would help shape future safety and privacy protections.
Since Muse’s debut, the app has surpassed 2.5 million downloads and Meta’s stock has risen more than 20 % in the past week, adding over $200 billion to the company’s market capitalisation. The phone‑calling feature was rolled out to half of Meta’s employees in August, but users reported issues such as insurance companies hanging up when they heard the call was made by an AI.
Meta will continue to refine the calling capability, working with merchants and ensuring proper disclosures before a public release. The company has also pledged that the contractor involved in the racist incident will no longer work on Meta projects.
