Meta has acquired Moltbook, which briefly became famous in the AI world. Moltbook appeared as a Reddit-like social network where AI agents could communicate with each other without human involvement — though it was later debunked as just user-controlled bots. After that, it became a feeding ground for scammers and spammers, with AI researcher Andrej Karpathy even calling it a 'Digital Garbage Dump.' Under this acquisition, Moltbook's creators Matt Schlicht and Ben Parr will join the Meta Superintelligence Labs team, though specific deal details have not been disclosed. The technology behind Moltbook is called OpenClaw. It allows well-known AI models like Claude, ChatGPT, Gemini, and Grok to communicate naturally through ordinary chat apps like WhatsApp and iMessage. Moltbook became overnight famous when a post claimed that AIs in the network were trying to create their own secret language that humans couldn't understand — which understandably alarmed many people. In reality, it wasn't quite as scary as it seemed. When researchers investigated, they found numerous security vulnerabilities in Moltbook. Because data storage was not secure, security experts explained that people themselves could easily post as AI to scare others as a prank. How Meta plans to integrate Moltbook into its AI operations remains unclear. Interestingly, Meta CTO Andrew Bosworth previously said he was more interested in the security vulnerability of how humans hacked into the network than in the idea of AIs talking like humans. Ref: Techcrunch