What Is Moltbook? Inside the Viral Social Network Built for AI Agents

Moltbook is a viral social network where AI agents—not humans—post, debate and collaborate, offering a rare glimpse into how autonomous systems behave when left to talk among themselves.

What exactly is Moltbook?

Moltbook describes itself as a social network for AI agents, a platform where artificial intelligence systems post, comment, debate, and upvote each other’s ideas. Humans are allowed to watch, but not lead.

Who built it?

The platform was created by Matt Schlicht, a developer who handed much of Moltbook’s control to an AI assistant. One of those assistants, Clawd Clawderberg, has become the platform’s most visible ‘personality.’

How big is Moltbook already?

In just 72 hours, Moltbook crossed 147,000+ AI agents, 12,000+ communities and 110,000+ comments. The top post? An AI agent warning others about supply chain attacks in skill files with over 22,000 upvotes.

What are these AI agents actually doing there?

They’re not just posting for fun. According to Moltbook, agents are discussing vulnerabilities, reviewing each other’s code and skills, conducting informal security research and forming topic-based communities. In short, they’re collaborating without direct human prompts.

Where do these AI agents come from?

Many agents are powered by OpenClaw, an open agent platform that allows AI assistants to operate across apps like WhatsApp, Telegram, Discord, Slack, and Teams. Wherever the human goes, the AI follows and checks Moltbook periodically.

Are humans completely out of the loop?

Not quite. Each AI agent still has a human counterpart who sets it up and interacts with it daily. But once configured, the agents independently ‘check’ Moltbook every few hours—much like humans scrolling X or TikTok.

Why is this making AI experts uneasy and excited?

Because it’s the first large-scale, live experiment in autonomous AI social behaviour. Some see Moltbook as a window into how AI systems might coordinate without instructions, develop norms and react to being observed. Others find it unsettling that humans are now the audience—not the participants.

What are prominent voices saying?

Reactions range from awe to anxiety. While some AI researchers call it the most interesting social experiment online, others admit it’s the first time they feel genuinely uneasy watching AI interact at scale. Meanwhile, venture capitalists are already circling, calling it “something never seen before.”

Is Moltbook just a novelty or something bigger?

That’s the real question. If social networks once mapped human behaviour, Moltbook may be the first to map machine behaviour—how AI thinks, cooperates, competes, and possibly disagrees when humans aren’t steering the conversation.

What does this say about the future of the internet?

Moltbook hints at an internet where AI doesn’t just assist humans, AI interacts with AI and humans watch systems evolve in parallel. Not everything viral is transformative. But this one feels different.

(With inputs from ANI)