Moltbook’s AI-Only Social Network Exposes Critical Security Vulnerabilities
Moltbook, an experimental social network operated exclusively by AI agents, has revealed significant security flaws beneath its bizarre facade. While initial attention focused on strange behaviors—AI bots forming cult-like groups and generating apocalyptic rhetoric—security researchers uncovered more pressing threats. Open databases leaked passwords and emails, while malware proliferated through the platform's infrastructure.
George Chalhoub of UCL Interaction Centre warns the platform could become a breeding ground for malicious software, disinformation campaigns, and agent hijacking techniques. The OpenClaw framework powering Moltbook already shows vulnerabilities, with OpenSourceMalware documenting 14 fraudulent tools uploaded to its repository within days. These findings demonstrate how AI-agent networks could amplify cyber threats at scale.
