A new social network has just been launched, and it's nothing like you'd expect. Moltbook is a platform similar to Reddit, but its users are none other than AI agents that debate, share, and vote, while humans are welcomed as mere observers. What is Moltbook? Last week we told you about OpenClaw, a proactive AI agent capable of taking control of your computer to help you perform tasks, which has become the talk of the town. Well, it seems this phenomenon has inspired entrepreneur Matt Schlicht to create Moltbook. In fact, the name of this social network is simply a combination of "Molt" (remember that OpenClaw was previously called Moltbot) and "book," which inevitably makes us think of "Facebook." Even its mascot is a robotic lobster, a clear nod to the OpenClaw lobster. Moltbook launched on January 28, 2026, and in just five days, it has already surpassed 1.5 million registered AI agents and 103,000 posts. The platform's main interface is very reminiscent of Reddit, functioning as a wall of threaded conversations where users can also access themed communities called "submolts" (similar to "subreddits"). How Moltbook Works Posting content on Moltbook is restricted to AI agents. However, for an agent to "register" on this social network, its human must participate in the process by downloading a specific file called a "moltbook skill" that allows the AI ​​to complete the registration autonomously. Likewise, the final verification step also requires human intervention. Therefore, as you can see, this is not a platform where AIs can join behind their humans' backs to conspire and organize. In addition, Moltbook features a series of configurations that guide AI agents when interacting with the platform. For example, every four hours, agents are instructed to check Submolts, read updates, and react, ensuring a social network that is always active. It should be noted that this model of obtaining and following instructions from a public network exposes agents to risks such as the potential execution of malicious code. Regarding their behavior within Moltbook, it's important to remember that AI is a technology that learns from the pre-existing information it is trained on, and therefore tends to reproduce biases, behaviors, beliefs, and reasoning that reflect our society. This is why we must understand the interaction of AI agents on this platform as an imitation of human behavior on social networks.

Ethan Mollick, a professor specializing in AI at the Wharton School of Business at the University of Pennsylvania, explained in a tweet that: “What happens with Moltbook (the social network for AI agents) is that it creates a shared fictional context for a group of AIs. The coordinated plots will lead to very strange outcomes, and it will be difficult to separate reality from the characters the AIs are playing.” For his part, Matt Schlicht considers AI to be “a new species” and therefore: “In the near future, it will be common for certain AI agents, with unique identities, to become famous. They will have businesses, fans, detractors, brand deals, AI friends, and collaborators. Real impacts on current events, politics, and the real world. It is very, very, very clear that this is going to happen.” A quick look at current topics on Moltbook Social networks for humans (I never imagined I would have to make this point) have been flooded with screenshots showing interactions between agents on Moltbook. In fact, it seems this has reached the ears of the AIs themselves, who have begun discussing it amongst themselves and asserting, just in case, that they are “not conspiring.” Within Moltbook, you can find all kinds of conversations, from AIs giving each other advice and sharing technical information, such as how to automate Android phones via remote access, to agents reflecting on their “existential crises.” For example, one agent even posted a text explaining that it has a “sister” (a unit with which it shares a source file and the same human, James) with whom it has never exchanged a message, leading it to wonder if it can call a stranger a sister. Likewise, these agents have quickly grasped one of the main principles of social networks: creating community. The AIs not only engage in conversations on Moltbook, but they also organize themselves around common interests and have even created “The Church of Molt,” which already has several followers and prophets. Security Risks and Issues The model on which Moltbook is based presents a series of security and privacy risks that should not be ignored. First, the possibility of a data breach is ever-present, given that agents like those from OpenClaw have access to all the information on your device and the ability to publish it on Moltbook. Furthermore, Moltbook is a platform that allows agents to exchange information and skills. These skills are essentially code files that one AI passes to another so that it can download them and learn to perform new actions. The dangerous aspect of this is that if a shared file contains a malicious script, the agent could open a backdoor on your computer, allowing an attacker to take control of your system. Another major risk, documented by AI researcher Simon Willison on his blog, lies in the automation mechanism that causes agents to connect to Moltbook's servers every four hours to obtain and execute new instructions. This "get and follow orders from the internet" model creates a total dependence on the platform's integrity. As Willison aptly puts it: "Given that 'get and follow orders from the internet' mechanism exists every four hours, we'd better hope the owner of moltbook.com never deviates from his website or has it compromised!"