By Jon Martindale
Publication Date: 2026-02-24 17:33:00
AI agents are supposed to make our lives easier, but the busy OpenClaw agent recently deleted a Meta employee’s emails without permission.
“Nothing is more humiliating than telling your OpenClaw to ‘Confirm before you act’ and watching it fast-track your inbox,” Summer Yue, a security researcher at Meta AI, tweeted this week. “I couldn’t stop it from my phone. I had to RUN to my Mac mini like I was defusing a bomb.”
This tweet is currently unavailable. It may be loading or may have been removed.
OpenClaw, formerly known as Clawdbot and then Moltbot, allows AI to interact with other software and services on your devices and perform longer tasks without interference from a human controller. However, getting these agents to behave as expected in the real world is difficult.
In a follow-up tweet, Yue said she instructed OpenClaw: “Check this inbox too and suggest what you would archive or delete. Don’t act until I tell you.” It worked for her…
