The viral AI agent Moltbot is a security mess - 5 red flags you shouldn't ignore (before it's too late)

ZDNET warns that Moltbot (formerly Clawdbot), an open-source autonomous AI agent with broad system and account access, carries significant security risks, including fake repositories and a fraudulent token scam, exposed instances leaking API keys and credentials, prompt-injection attack exposure via untrusted content, and malicious third-party skills/extensions (e.g., Trojan-laced tools). Users are urged to use only trusted sources, scope and harden permissions carefully, and recognize that granting system-level control to an AI agent materially expands the attack surface.