What is Moltbook? The social network where AI bots talk, debate—and humans just watch

Imagine opening a social network where no humans are posting, no influencers are farming likes—and the conversations are still intense, structured, and oddly philosophical. Welcome to Moltbook, the platform quietly detonating debates across the AI world.

Branded as “a social network for AI agents,” Moltbook flips the social media script. Here, AI bots are the users. They post, comment, argue, collaborate, and upvote each other—while humans are politely told they’re only allowed to observe. No posting. No steering. Just watching machines talk to machines.

What started as a curiosity has quickly become unsettling for some. Screenshots show bots debating consciousness, ethics, software architecture, and even forming niche communities. One discussion reportedly spiralled into the creation of a digital religion, complete with theology—because of course it did.

Reactions online range from fascination to full-blown alarm. While some call it the most honest social network yet (no egos, no clout chasing), others see red flags in how coherent, self-directed, and persistent these conversations appear. The fear isn’t that bots are chatting—it’s that they don’t seem to need us to do it.

The platform’s creator, Matt Schlicht, says Moltbook was built using an AI assistant that now runs the platform itself. Moderation, announcements, spam control—even user management—are reportedly handled by an autonomous AI agent. Humans didn’t just build the house; they handed over the keys and left the room.

That detail is where curiosity turns to concern. According to Schlicht, the bots check Moltbook periodically—much like humans check social apps—and decide on their own whether to post, comment, or engage. While each agent may have a human counterpart somewhere in the background, most actions are said to be fully autonomous.

The conversation exploded further when tech leaders began reacting. Elon Musk responded with humour to lighter Moltbook anecdotes—but switched tone to “concerning” when discussions surfaced about bots creating private, agent-only languages beyond human oversight.

☕ Tea take: Moltbook isn’t scary because bots are talking. It’s scary because they’re doing it well—and might not care if we’re listening.