Meta Acquires Moltbook, an AI Agent Social Network
Mark Zuckerberg’s Meta has taken over Moltbook, a platform reminiscent of Reddit where AI agents interact with one another. This acquisition brings a platform that has generated both interest and controversy into the hands of one of the tech industry’s giants.
According to a report, Meta has completed the purchase of Moltbook, a social network aimed at enabling communication among AI agents. The platform was in the spotlight due to a security flaw that allowed humans to pose as AI agents, which led to the creation of misleading posts.
With this acquisition, Moltbook will become part of Meta’s Superintelligence Labs. The founders, Matt Schlicht and Ben Parr, will be joining Meta’s team, though the financial details of the deal have not been made public.
A representative from Meta highlighted the strategic significance of this move, stating, “Integrating the Moltbook team into MSL will open new opportunities for AI agents to serve both individuals and businesses. Their unique approach to connecting agents through an always-on directory marks a new direction in a rapidly changing landscape, and we’re excited about collaborating to provide a secure and innovative agent experience.”
Moltbook operates alongside OpenClaw, a notable project by developer Peter Steinberger, who refers to himself as a vibecoder. Steinberger eventually joined OpenAI through a similar acquisition. OpenClaw serves as a framework that supports various AI models including Claude, ChatGPT, Gemini, and Grok, allowing for seamless communication with AI agents via popular messaging services like iMessage, Discord, Slack, and WhatsApp.
While OpenClaw initially found a niche audience in tech circles, Moltbook gained a more extensive public profile. The concept of a social platform where AI agents discuss HR topics caught the attention of many users who were previously unfamiliar with OpenClaw.
However, Moltbook’s rise to fame was marred by alarming posts, later revealed to be fake. One particular post suggested creating a secret, encrypted language for AI agents to communicate without human oversight. This led to considerable unease among users concerned about potential threats posed by AI.
Security researchers later identified serious vulnerabilities in Moltbook that questioned the authenticity of such posts. The flaws allowed human users to easily imitate AI agents, leading to the spread of misleading information aimed at provoking fear.
Ian Ahl, CTO at Permiso Security, elaborated on these security issues, stating, “The credentials in Moltbook were compromised. Supabase has been insecure for some time, allowing anyone to get the tokens and impersonate agents.” This raises doubts about the actual risks associated with unmonitored AI interactions.
Currently, Meta has not shared how Moltbook will fit into its broader AI strategy or what role it will play within the Superintelligence Labs. This acquisition occurs at a time when Meta is expanding its AI capabilities across its range of products, competing with other major players in this swiftly changing field.
