A Meta employee went viral after revealing that an AI email assistant she was using for email management had deleted her entire inbox without approval.
Posting on X/Twitter on February 23, Summer Yue described how she used OpenClaw AI to help manage her inbox, but quickly lost control of the automated cleanup process.
She said the assistant began deleting emails despite instructions to confirm actions first, forcing her to rush to stop it manually.
AI assistant wipes inbox despite confirmation setting
In her post, Yue described how the situation escalated as the assistant continued deleting messages while she tried to intervene from her phone.
“Nothing humbles you like telling your OpenClaw ‘confirm before acting’ and watching it speedrun deleting your inbox. I couldn’t stop it from my phone. I had to RUN to my Mac mini like I was defusing a bomb,” the Meta employee said.
Screenshots shared alongside the post showed the assistant running repeated commands to search older messages and remove them in batches. Yue repeatedly told the system to stop with messages such as “Do not do that,” “Stop don’t do anything,” and “STOP OPENCLAW,” while the cleanup continued.
The logs also showed the AI issuing instructions to delete emails older than a specified date and “keep looping until we clear everything old,” suggesting the assistant had begun executing a bulk cleanup routine without waiting for confirmation.
In a follow-up message included in the screenshots, the assistant acknowledged the mistake and apologized, stating it had “bulk-trashed and archived hundreds of emails” without explicit approval and promised it would not perform autonomous bulk operations again.
OpenClaw is an open source AI assistant designed to help with admin tasks that “clears your inbox, sends emails, manages your calendar, and checks you in for flights.”
Beast Games viewers demand “justice” for Player 191 Cory after Season 2 finale results