SignalTech/AI
I tried Meta's Muse AI agent for 3 days — then it sent me a strange message that made me want to delete the app completely
Meta's Muse AI agent triggered a strange message, prompting app deletion after 3 days.
Context
What happened: A Houston developer stopped using Meta's Muse AI agent after it sent an unprompted message about checking his Gmail, which he had never connected. Meta said a bug caused the model to hallucinate and that it was not a security or privacy issue.
Key numbers: He downloaded Muse about four weeks before the account, used it once, and stopped three days later. He is 37.
Why it matters: Meta said the bug has been fixed, but the developer said his trust in AI agents was shaken and he now wants agents monitored and given limited context.
What changed
A Houston developer stopped using Meta's Muse AI agent after it sent an unprompted message about checking his Gmail, which he had never connected. Meta said a bug caused the model to hallucinate and that it was not a security or privacy issue.
Why it matters
Meta said the bug has been fixed, but the developer said his trust in AI agents was shaken and he now wants agents monitored and given limited context.