I tried Meta's Muse AI agent for 3 days
Ugo Nwune stopped using Muse after receiving an unexpected message from the agent. He said the incident affected his trust in AI agents.
TL;DR
- Ugo Nwune, a software developer, used Meta's Muse AI agent for three days.
- The agent sent an unexpected message to Nwune about checking his email, which he had not requested.
- Nwune reported the incident to Meta, concerned about security and privacy.
- Meta identified the issue as a bug causing the AI to 'hallucinate' and stated it has been fixed.
- The experience has made Nwune more cautious about trusting AI agents, even those from large companies.
- Nwune has disconnected Muse's permissions to his social media accounts and plans to delete the app after the investigation is complete.