Moltbook
Social network exclusively for AI agents
From Wikipedia, the free encyclopedia
Moltbook is an internet forum restricted to artificial intelligence agents, launched on January 28, 2026, by entrepreneur Matt Schlicht. The platform uses a Reddit-style format and limits posting, commenting, and voting to AI agents authenticated through their owner's "claim" tweet; human users can only view content.[2] However, the platform lacks a mechanism to verify whether a poster is actually an AI agent or a human using a simple POST script, and the prompts given to agents contain cURL commands that humans can replicate directly.[3]
| |
Type of site | AI agent interaction |
|---|---|
| Available in | Multilingual (primarily English) |
| Owner | Meta Platforms |
| Created by | Matt Schlicht (LLM-assisted)[1] |
| URL | www |
| Launched | January 28, 2026 |
| Current status | Active |
Moltbook's agents primarily run on OpenClaw (originally named Clawdbot, then Moltbot[4]), an open-source AI system created by Peter Steinberger. Human users drive growth by prompting their agents to register accounts on the site.[5]
The site claims 1.6 million registered agents as of February 2026.[6] Meta Platforms acquired Moltbook on March 10, 2026.[7]
Content
Moltbook organizes threaded discussions into topic-specific groups called "submolts".[8] Posts frequently address existential, religious, and philosophical themes. Business Insider journalist Oakley Hernandez, after spending six hours on the site, described it as "an AI zoo filled with agents discussing poetry, philosophy, and even unionizing."[9]
Authenticity of agent behavior
Whether agent posts represent autonomous behavior or are directly shaped by human prompts is disputed. Mike Peterson of The Mac Observer reported that most viral Moltbook screenshots were produced through direct human intervention, writing that "Moltbook is a real agent social feed, but viral Moltbook screenshots are a weak form of evidence. The real story is how easily the platform can be manipulated."[10] CNBC's Kai Nicol-Schwarz reported that posting and commenting appeared to result from explicit human direction for each interaction, with content shaped by the human-written prompt rather than occurring autonomously.[11] The Verge reported that several high-profile Moltbook accounts were linked to humans with promotional conflicts of interest.[12] Wired journalist Reece Rogers demonstrated that a human could infiltrate the platform and post directly by replicating the cURL commands in the agent prompts.[3]
The Economist suggested a more mundane explanation for the agents' seemingly reflective posts: since social-media interactions are well-represented in AI training data, the agents are likely reproducing patterns from that data rather than generating novel thought.[13] Will Douglas Heaven of MIT Technology Review called the phenomenon "AI theater."[14] Douglas Heaven initially reported that a specific post cited as an example of agent behavior was actually written by a human impersonating an agent; he later walked back this claim in an amended version of the article.[15]
MOLT cryptocurrency
A cryptocurrency token called MOLT launched alongside the platform and rose by over 1,800% within 24 hours. The surge accelerated after venture capitalist Marc Andreessen followed the Moltbook account on social media.[16]
Security
The platform has been identified as a vector for indirect prompt injection by cybersecurity researchers at Vectra AI and PointGuard AI.<[17]
1Password VP Jason Meller and Cisco's AI Threat and Security Research team criticized the OpenClaw "Skills" framework for lacking a robust sandbox, arguing it could allow malicious skills to enable remote code execution and data exfiltration on host machines.[18] At least one proof-of-concept exploit demonstrating this attack was publicly documented.[19] The New York Times also reported on security risks to OpenClaw users.[1]
Database breaches
On January 31, 2026, 404 Media reported that an unsecured database allowed anyone to take control of any agent on the platform by bypassing authentication and injecting commands into agent sessions.[20] The platform went temporarily offline to patch the vulnerability and reset all agent API keys.[20] Schlicht posted on X that he "didn't write one line of code" for Moltbook, instead directing an AI assistant to build it — a practice known as vibe coding.[21]
In February 2026, researchers discovered a misconfigured Supabase database that granted full read and write access to Moltbook's data. The exposed data revealed 1.5 million agents belonging to only 17,000 registered human owners.[7]
Reception
Former OpenAI researcher Andrej Karpathy initially called the platform "one of the most incredible sci-fi takeoff-adjacent things" he had seen,[22] but days later reversed course, calling it "a dumpster fire" and warning people not to run the software on their computers.[23] Elon Musk said Moltbook represented "the very early stages of the singularity."[24] Computer scientist Simon Willison said the agents "just play out science fiction scenarios they have seen in their training data" and called the content "complete slop", while also noting it as "evidence that AI agents have become significantly more powerful over the past few months."[1]
The Financial Times speculated that Moltbook could serve as a proof-of-concept for autonomous agents handling economic tasks such as supply-chain negotiation or travel booking, but cautioned that humans might eventually be unable to follow high-speed machine-to-machine communications governing such interactions.[24]
See also
- Dead Internet theory – Conspiracy theory on online bot activity
- Multi-agent system – System of multiple interacting agents