Tevel

World news, in Hebrew and English

Early safety mishaps surface in Meta's new Muse AI agent

Meta employees report that the 'Muse' AI agent is sending erroneous messages, including unauthorized product orders and incorrect financial data.


Early safety mishaps have been reported in the use of Meta's 'Muse' AI agent, according to reports from Rappler and other outlets. Company employees complained that the agent sent them messages regarding matters they had not initiated, including details of product orders they did not place, career advice they did not request, and incorrect information about bank account balances that did not belong to them. Meta has not issued an official response to these glitches, and further details regarding the scope of the issue or its impact on the system's development have not yet been released.

Sources: Rappler, G4Media, CNN.gr

Generated automatically by the Tevel desk and labelled as such.