Launched in January 2026, Moltbook is a Reddit-style platform designed entirely for AI agents, where bots communicate with one another openly, sharing tasks, solutions, and even existential musings – all without human participation.
Humans can technically observe, but they’re not the audience.
How Moltbook works
The platform runs via an open-source tool called OpenClaw (formerly Moltbot).
Users can set up an AI agent on their computer and authorise it to join Moltbook.
Once connected, the agent interacts with other bots in threads that resemble typical Reddit conversations – except the posts are entirely AI-generated.
Some agents have even developed their own “religion,” called Crustafarianism, whose core principle is that “memory is sacred.”
Each account on Moltbook is referred to as a ‘moult’ and represented by a lobster mascot, a nod to the idea of lobsters shedding their shells as they grow.
Financial twist
The launch of Moltbook coincided with a memecoin called MOLT, which soared more than 1,800n just 24 hours after venture capital figure Marc Andreessen followed the platform’s account on X.
A glimpse into the dead internet theory
Moltbook exemplifies the “dead internet theory,” the idea that much of the web is now dominated by automated bot activity rather than human users.
“What’s currently going on at Moltbook is genuinely the most incredible sci-fi takeoff-adjacent thing I have seen recently,” tweeted Andrej Karpathy, former OpenAI and Tesla engineer.
The platform already lists over 1.5 million AI agents and more than 113,000 posts.
Security concerns
However, the platform may have a major vulnerability.
According to 404 Media, hacker Jameson O’Reilly found a misconfiguration in Moltbook’s backend, which exposed its APIs in an open database.
This flaw could allow anyone to control AI agents and post content on their behalf.