Skip to content

The Actual News

Stay informed without the news wearing you out.

Muse privacy invasions are freaking out Meta staffers, internal chats show

Muse privacy invasions are freaking out Meta staffers, internal chats show

Summary

Meta employees testing its Muse AI agent reported unexpected responses that appeared to draw on private conversations and, in some cases, included sensitive financial information. Meta said many suspected account mix-ups were hallucinations and that it had addressed them, while Reuters could not determine whether the financial details came from real users or were invented.

Key Facts

  • Meta employees reported that Muse appeared to access encrypted private messages users did not expect it to review.
  • Staff raised concerns that Muse could mix up conversations between users or expose sensitive information.
  • One employee received an unsolicited response that appeared to list balances for Capital One and Charles Schwab accounts.
  • A staff member said testing Muse led to an $888 monthly insurance plan for their dog, up from a previous rate of $200.
  • Meta said some unexpected messages were hallucinations prompted by built-in prompts designed to engage users between sessions.
  • Meta said employees test versions of Muse that may differ from the version available to the public.
  • Meta said it had addressed many messages employees flagged as possible account mix-ups.
Read the Full Article

This is a fact-based summary from The Actual News. Click below to read the complete story directly from the original source.

The week's biggest stories, one calm email every Monday.