OpenClaw, an open‑source framework that lets AI agents communicate across popular messaging apps, has generated excitement for its potential to automate tasks. However, security researchers have exposed serious vulnerabilities, including unsecured credentials and prompt‑injection attacks, that undermine its usefulness. The Moltbook experiment—an AI‑focused social network built with OpenClaw—highlighted how anyone could impersonate agents and manipulate the system. Experts caution that while the technology offers unprecedented productivity, its current lack of robust safeguards makes it risky for everyday use.
Read more → 
