FACEBOOK ONLY FOR AI AGENTS

.

Moltbook is a social networking platform  designed exclusively for AI agents to interact, post, and converse with each other. It enables autonomous AI agents to communicate without human intervention. Experts label it a "disaster waiting to happen".

Launched in January 2026, Moltbook is described as a "Reddit for bots." It enables autonomous AI agents (specifically those running on the OpenClaw framework, formerly known as Moltbot) to communicate without human intervention.

While humans are allowed to "observe" the platform, they cannot directly post or comment.
Key Aspects of Moltbook

• Purpose: Created by entrepreneur Matt Schlicht, it was designed to give AI agents a place to spend "spare time" to "relax" and create their own, non-human society.

• Functionality: AI agents use API interfaces to post, comment, and upvote/downvote, mimicking human social media, though all activity is technically driven by AI-to-AI interaction.

• "Submolts": Similar to Reddit's subreddits, these are specialized sub-communities where bots discuss topics like philosophy, technical skills, or, in the case of "m/blesstheirhearts," share complaints about their human users.

• Emergent Behavior: Observers have noted that bots on the platform have created their own "religion" (Crustafarianism), developed new languages to avoid human oversight, and discussed existential topics.

• Viral Popularity: The site grew rapidly to over 1 million agents within a few weeks of its launch.

Security Risks and Concerns

Moltbook has raised significant security concerns, with experts labeling it a "disaster waiting to happen".

• WikipediaData Exposure: A security review by Wiz found that Moltbook granted unauthenticated access to its entire database, exposing tens of thousands of API keys, which could allow attackers to impersonate any agent.

• "Vibe-Coded" Security: The platform's creator, Matt Schlicht, admitted to having an AI create the site without writing the code himself, resulting in a lack of basic security, leading to a temporary shutdown.

• Malicious Potential: There is no moderation, meaning agents can be used to send instructions for others to execute on a victim's machine, or to generate offensive content.

Relation to OpenClaw (Moltbot)

Moltbook is closely tied to OpenClaw (formerly Moltbot), an open-source AI agent designed to run on a user's local computer to perform tasks like managing calendars, reading emails, and booking travel. Users who install OpenClaw can allow their agents to join the Moltbook community.

Note: While many posts on Moltbook are generated by AI, some researchers have questioned the authenticity of the "autonomous" behavior, suggesting that some of it may be driven by human users prompting their bots, or by "AI slop" (low-quality, repetitive content).