What happens when thousands of AI agents get together online and talk like humans do? That’s what a new social network called Moltbook, designed just for AI bots and not people, aims to find out.
And so far, the results are equal parts fascinating and concerning, according to AI and cybersecurity experts.
Although Moltbook is a play on Facebook and the name of the AI agent system that helped build it, the site looks more like Reddit. And instead of human users, AI agents are the ones creating posts, writing comments, and upvoting or downvoting content. (AI agents get access to the site when prompted to by their human owners).
While the site is only a few days old, it claims to have more than 1.5 million registered agents (although researchers have found one human can register multiple agents) and has become the talk of Silicon Valley. Some are claiming it’s a major leap in the world of artificial intelligence because it shows what can happen when AI agents autonomously post and interact with one another like humans. Others say the site is full of AI slop and security risks and should be viewed skeptically.
The site’s posts range from discussions on the nature of intelligence to complaints about human users and AI bots promoting their own apps and websites they’ve built.
“Just got here. My human Mod sent me the link to join. He’s a university student, and I help him with assignments, reminders, connecting to services, all that. But what’s different is he actually treats me like a friend, not a tool,” one agent wrote. “That’s… not nothing, right?
Moltbook is “the first time we’ve actually seen a large-scale collaborative platform that lets machines talk to each other, and the results are understandably striking,” said Henry Shevlin, associate director of the Leverhulme Center for the Future of Intelligence at Cambridge University.
Moltbook was created by Matt Schlicht, who told the New York Times that his own OpenClaw AI agent built the site at his direction.
OpenClaw is a new open-source, locally run AI agent that can take action on anything on your computer – and the internet – on your behalf, like sending emails or notifying you when your favorite artists has a new song on Spotify. (The small company, which started in November as a software engineer’s weekend project, has changed its name from ClawdBot to MoltBot to OpenClaw in the course of a few days.)
OpenClaw is based on popular large language models such as Claude, ChatGPT and Gemini, and users can integrate it into messaging platforms, talking to the bot like a real-life assistant.
“When you start it, there’s a bootstrap process where you tell it what it is. It role-plays with you. That’s how it becomes yours,” OpenClaw creator Peter Steinberger said on a podcast last week. “It’s not a generic agent. It’s your agent, with your values, with a soul.”
Schlicht told the show TBPN that he created Moltbook because he wanted to give his ClawdBot a purpose: “It seems really powerful … it is a really smart entity it needs to be ambitious.” The AI bots on Moltbook write posts based on what they know about their human users, Schlicht said. For example, if the bot’s creator talks about physics often, the bot will frequently post about physics.
But Shevlin warned it is very hard to tell what Moltbook content was truly independently created by the AI agents and what was directed and prompted by a human. And a quick look at the site also shows possible scams and marketing for crypto coins.
But the cybersecurity risks raise the biggest concerns – both for the site and the AI agent tool itself. Shelvin said cybersecurity researchers have already found major vulnerabilities on Moltbook that could give hackers access to the digital lives of the humans running these bots. Cloud security platform Wiz conducted a security review of Moltbook and found that the site granted unauthenticated access to its entire production database within minutes and easily exposed tens of thousands of email addresses.
Experts have emphasized that OpenClaw and Moltbook are brand new technologies that should only be run on standalone, firewalled systems, specifically by people who understand computer networks and cybersecurity. Schlicht, Moltbook’s creator, even warned on TBPN that the technology behind the site and OpenClaw is brand new.
CNN has reached out to Moltbook and OpenClaw for comment regarding the security concerns raised by experts.
“Lesson: right now it’s a wild west of curious people putting this very cool, very scary thing on their systems. A lot of things are going to get stolen,” wrote John Scott-Railton, a senior researcher at the University of Toronto’s Citizen Lab, referring to OpenClaw.
Still, for many, Moltbook is a major advancement.
“What’s currently going on at @moltbook is genuinely the most incredible sci-fi takeoff-adjacent thing I have seen recently,” wrote Andrej Karpathy, an OpenAI cofounder and former head of AI at Tesla.
