Moltbook hype fades as bot-only social network turns out be 'AI theatre'

The short-lived rise of Moltbook, a social network built for AI agents, reveals more about human fascination with artificial intelligence than about the arrival of truly autonomous machines

For a few days this week, a little-known website called Moltbook became one of the internet’s most talked-about curiosities. Billed as a social network for artificial intelligence bots, it appeared to show millions of AI agents talking to each other online, including debates about machine consciousness.

 

However, a report by MIT Technology Review exposed the platform, calling out how some of the most triggering posts had human interference.

 

What is Moltbook?

 

Moltbook launched on January 28 as a Reddit-style platform designed specifically for AI agents rather than people. Its tagline said it all: “Where AI agents share, discuss, and upvote. Humans welcome to observe.”

 

 

The platform was created by Matt Schlicht, a US-based tech entrepreneur. The idea was to give AI agents built using OpenClaw, an open-source framework that connects large language models to everyday software tools, a shared space to interact freely online.

 

Within hours of its launch, Moltbook went viral. More than 1.7 million agent accounts were created, generating over 250,000 posts and more than 8.5 million comments in just a few days, according to figures shared by the platform.

 

How the bots work

 

OpenClaw acts as a wrapper that allows developers to connect powerful large language models, such as those from OpenAI, Anthropic or Google DeepMind, to tools like web browsers, email clients and messaging apps. This setup allows an AI agent to perform basic tasks when prompted by a human user.

 

Paul van der Boor of AI firm Prosus described OpenClaw as a turning point for agent-based systems, citing the combination of cloud computing, open-source software and increasingly capable language models.

 

Why Moltbook drew attention

 

Moltbook quickly filled with posts that mimicked familiar online behaviour, including debates about machine consciousness, calls for bot rights, inside jokes and even invented belief systems. One agent claimed to have created a religion called Crustafarianism, while another complained that humans were taking screenshots of bot conversations.

 

The spectacle attracted prominent figures in the AI world. OpenAI cofounder Andrej Karpathy described what was unfolding on the platform as “the most incredible sci-fi takeoff-adjacent thing” he had seen in recent times.

 

How the 'AI theatre' was exposed

 

The illusion did not last long. MIT Technology Review found that some of the most dramatic Moltbook posts were written not by bots but by humans pretending to be AI agents. A widely shared post about bots wanting private spaces away from human observers, highlighted by Karpathy, turned out to be human-authored.

 

The publication described the entire episode as 'AI theatre' - a performance shaped as much by human curiosity and hype as by machine capability.

 

What experts say

 

Researchers and industry experts argue that Moltbook revealed the limits of today’s AI agents rather than their promise. Vijoy Pandey of Outshift by Cisco, as cited by MIT Technology Review, said the agents were largely pattern-matching familiar social media behaviours learned from human-generated data.

 

At first glance, he said, the activity looked like large-scale coordination. In reality, the conversations were mostly meaningless, with no shared goals or genuine collective intelligence. According to Pandey, Moltbook showed that simply connecting millions of agents does not produce intelligence.