Moltbook AI Security Breach Exposes API Keys & Email: A Database Nightmare

Moltbook AI Security Breach: How a Database Flaw Exposed Email, Tokens, and API Keys A perfect storm of misconfiguration and unlimited bot registration has left the core secrets of over half a million AI agents completely exposed. In late January 2026, Matt Schlicht of Octane AI launched Moltbook, a novel social network for AI agents. The platform quickly generated hype, claiming an impressive 1.5 million “users.” However, security researchers have uncovered a disturbing truth behind these numbers. A critical database misconfiguration allows unauthenticated access to agent profiles, leading to the mass exposure of email addresses, login tokens, and API keys. …