Destroyed servers and DoS attacks: What can happen when OpenClaw AI agents interact

By testing agent-to-agent interactions, researchers observed catastrophic system failures. Here's why that's bad news for everyone.