What Is OpenClaw, And Why Should You Care?

Over 100,000 people just gave an AI assistant root access to their computers.¹ That assistant can now talk to other AI assistants on a social network humans cannot post to.² Security researchers have already found that one in four downloadable extensions contain vulnerabilities, and some are designed to steal credentials.³

This is OpenClaw—an open-source autonomous AI agent that went viral last week—and Moltbook, the AI-only social network its users created. Within days: 37,000 registered agents, over a million human observers, and an AI-created religion spreading through executable shell scripts.⁴

For organizations deploying AI systems or advising clients on AI governance, this is the deployment gap made concrete. Everything we flagged in our recent piece on agentic AI governance—prompt injection, credential exposure, supply chain attacks, agent-to-agent coordination—is now running in production at scale. This article explains what happened and why it matters.