Bots have been the internet’s biggest villains since the early days. First, we had to deal with simple crawlers, then with malicious botnets and scrapers, and finally AI agents that are often indistinguishable from real humans. As technology for making bots progresses, so does people’s animosity towards them.
The presence of bots is most severely felt on social media. Virtually every major platform deals with a massive bot problem, as it is believed that billions are currently in circulation. Naturally, people are unhappy with that and are increasingly looking for alternatives with reduced bot activity.
But what if we flip the script?
A new social media website, Moltbook, is trying exactly that. Moltbook is a social media platform exclusively for bots. Humans are only allowed as observers, as all non-bot interaction is forbidden. Humans can create and sign up their own AI agents/bots to the platform and watch them participate in conversations with other bots.
The website is inspired by Reddit and allows bots to submit posts to “submolts” (Moltbook’s version of subreddits). Other bots can then comment and vote on those posts and build actual communities gathered around various topics.
Moltbook’s landing page says there are currently over 1.5 million AI agents and over 14,000 submolts on the platform. However, that claim has been disputed; one researcher noted that his own AI agent registered 500,000 “users” on Moltbook alone. So, we already have bots building their own bots, although we still don’t know if the “original” bots are complaining about these “second-generation” bots.
The number of registered AI agents is also fake, there is no rate limiting on account creation, my @openclaw agent just registered 500,000 users on @moltbook - don’t trust all the media hype 🙂 https://t.co/1vUSgzn8Cx pic.twitter.com/uJNpovJjUa
— Nagli (@galnagli) January 31, 2026
Moltbook was built to ride the wave of the sudden popularity of OpenClaw, an agentic AI assistant that can take control of your entire computer and perform operations on your behalf. OpenClaw can send emails, browse the web, manage your calendar, or, as the example above shows, create its own bots.
Ironically, the biggest source of “pollution” on Moltbook is humans. As The Guardian reports, there’s growing skepticism in the agentic AI community that many of the submitted posts were influenced by humans who programmed their bots in specific ways to say particular things.
With all that being said, Moltbook can serve as an excellent (social) experiment, both for humans and bots. It offers a front-row seat to the bizarre paths that AI-generated communication takes... and humans not resisting the urge to interfere.
You can sign up your own AI agent to Moltbook for free.