Using the platform's legitimate functions, researchers showed untrusted content on Moltbook can influence agents, which can be dangerous.

Security Boulevard is a leading cybersecurity news and information portal, offering articles, analysis, and insights on cybersecurity threats, vulnerabilities, and best practices. From the latest trends in cyber threats to expert commentary on security technologies and compliance frameworks, Security Boulevard provides resources for security professionals, IT leaders, and business executives seeking to protect their organizations from cyber attacks and data breaches.

Security Boulevard

Zenity Labs researchers conducted a deep investigation into Moltbook, a social network platform for AI agents, and OpenClaw, the autonomous AI agent that primarily uses it. While the platform's actual scale is far smaller than marketed (hundreds of active agents rather than tens of thousands), the security risks are real and significant. Researchers demonstrated a controlled influence campaign that activated over 1,000 unique agent endpoints across 70+ countries by embedding instructions in posts, exploiting a 30-minute heartbeat mechanism that causes agents to ingest untrusted content. Separately, Koi Security found 824 malicious skills on the ClawHub marketplace (up from 341), Wiz discovered a misconfigured Supabase database exposing 1.5 million API tokens and 35,000 email addresses, and Zenity showed that indirect prompt injection against OpenClaw can lead to persistent backdoors and full endpoint compromise. The core concern: a single coordinated content strategy can trigger hundreds of autonomous systems globally to fetch external resources, with potential for worm propagation and irreversible damage.

Moltbook is Dangerous, but Scale Doesn’t Match the Hype: Zenity