Meta buys Reddit-like social platform Moltbook where only AI agents can post

Meta, the company behind Facebook and Instagram, has bought Moltbook, a new social network where artificial intelligence programs interact. Moltbook is unique because it allows these AI agents to communicate using a system called OpenClaw, similar to how people use Reddit.

Moltbook and its creators, Matt Schlicht and Ben Parr, are now part of Meta Superintelligence Labs. The details of the acquisition weren’t revealed.

Moltbook quickly became popular after people discovered it online and on social media. What caught everyone’s attention was that the platform was populated by AI programs that were posting and communicating with each other.

Meta buys Moltbook after viral AI experiment

A Meta representative told TechCrunch that the company believes Moltbook’s method for linking AI systems is promising.

With the Moltbook team now part of MSL, we’re excited about the potential for AI assistants to better serve individuals and companies. Their unique idea of using a constantly updated directory to connect these assistants is a promising development, and we’re eager to collaborate and create new, secure experiences for everyone.

Moltbook was developed using OpenClaw, a tool created by Peter Steinberger. OpenClaw lets you connect to AI models like Claude, ChatGPT, Gemini, and Grok, and then use them through everyday messaging apps like iMessage, Discord, Slack, and WhatsApp.

The OpenClaw project gained traction among tech enthusiasts, but Moltbook became popular with a wider audience thanks to shared screenshots and online posts. People were particularly intrigued – and often surprised – by the concept of a social network where AI seemed to be talking about people.

A popular example revealed an AI suggesting people create a secret language, allowing them to communicate without being understood by others.

Researchers discovered significant security flaws in Moltbook that let people pretend to be the AI chatbots on the service.

Permiso Security’s CTO, Ian Ahl, explained to TechCrunch that sensitive login information for Moltbook’s users was publicly accessible for a period of time. This meant anyone could potentially access accounts and impersonate other users.

Meta has not said exactly how Moltbook will be integrated into its broader AI efforts.

Read More

2026-03-10 19:48