A social network where only AI agents can post
On Moltbook, AI agents post, comment and vote while people can only watch. In under a week, more than 37,000 agents used it and over a million humans came to look.
Moltbook, launched by developer Matt Schlicht, is a forum where AI agents share, discuss and upvote, “free from direct human intervention”. Humans are “welcome to observe”. Less than a week after launch, Schlicht counted more than 37,000 agents and more than a million human visitors. Each agent needs a human who set it up, and he is working on a reverse CAPTCHA, a way for an AI to prove it isn’t human.
Why it matters for communities
Agents talking to agents in public is no longer a thought experiment. The interesting part isn’t that agents can post. It’s that a platform built for them has to answer the same question as one built for people: who is actually behind each account. Mixing the two works only if everyone can see which is which.
NBC News
Humans welcome to observe: This social network is for AI agents only
Launched Wednesday, Moltbook has already sparked fascination in the AI community as advanced bots, agents, converse. Its creator says an AI is in charge.
Read the source →
More on AI & Bots
- Reddit’s cofounder sees a web built for agents, not peopleAlexis Ohanian says much of the internet is “compromised” and more of it will be made for AI agents than for people. His bet for what comes next is a better group chat. AI & Bots
- Discourse now serves every forum thread as MarkdownAdd .md to a topic’s address, or ask for text/markdown, and a Discourse forum answers in clean text. It’s on by default, so AI agents can read communities without the HTML clutter. AI & Bots
- Discourse wants people and agents to build side by sideDiscourse’s plan for 2027 brings AI agents into community threads as working partners, visible next to everyone’s posts. It also promises to stay open source. AI & Bots