MIT just audited 30 major AI agents. The findings are... sobering.

MIT just audited 30 major AI agents. The findings are... sobering.

Half have no published safety framework. One in three have no documented way to shut them down. Most don't even identify themselves as bots.

These are the tools companies are handing their email, their CRM, their customer data.

The takeaway isn't "avoid AI agents" — that ship has sailed. It's: what's your kill switch? Who's monitoring what they're actually doing?

Like living in tornado alley: you don't pretend tornadoes don't exist. You have a plan, know your safe place, and stay weather-aware.

Same deal with AI agents.