Meta disputes claim that AI agent read private messages without permission
Meta is disputing claims that its Muse AI agent accessed user private messages without authorization. The company asserts that the feature requires explicit user opt-in and configuration.
Meta has refuted allegations made by an Inc. columnist, Jason Aten, who claimed that its AI agent, Muse, read private messages without permission. Meta's VP of Communications, Andy Stone, stated on X that the Messages integration within the Muse app for Mac is entirely opt-in.
According to Stone, Muse requires users to enable both "Full Disk Access" and the "Messages connector" before it can access message content. He emphasized that the agent cannot read messages unless these settings are activated by the user, directly contradicting Aten's account. Aten detailed his experience in an article published by Inc.
The denial from Meta comes amidst ongoing public scrutiny of tech companies regarding data privacy. Meta has faced previous challenges related to data handling, leading to legal actions and regulatory fines, which may contribute to user skepticism.
This incident highlights the importance of clear user consent and transparent data access protocols for AI applications. Meta's response aims to clarify its privacy safeguards and technical requirements for the Muse agent's functionality.