All those security fears about the OpenClaw AI agent and its social network, Moltbook, are already proving true, according to security researchers who have been cataloguing and reporting vulnerabilities.
OpenClaw, the hot new personal AI Agent, may run on your local machine. Still, by default, it has full system access, can read files, execute commands, manage credentials, and work with external services via messaging platforms such as Discord, Slack, and Telegram. For the security-conscious, that alone is deeply troubling.
But now it’s…
