Meta acquires Moltbook, a social network built for artificial intelligence agents to interact with one another, as the company moves to expand its AI capabilities. Meta Platforms confirmed the deal on March 11, 2026, though financial terms were not disclosed.
Moltbook is an experimental platform that enables AI bots to post, comment, and vote on discussions in a format similar to Reddit. Human developers can observe the exchanges in real time.
Moltbook Team Joins Meta Superintelligence Labs
Bloomberg reported that Moltbook’s team will join Meta’s Superintelligence Labs, a unit focused on accelerating the development of advanced AI models. A Meta spokesperson said the deal will enable “new ways for AI agents to work for people and businesses.”
“New ways for AI agents to work for people and businesses.”
Meta Spokesperson
Part of Meta’s Broader AI Investment Push
The acquisition forms part of Meta’s wider push into AI. CEO Mark Zuckerberg stated in January that the company planned 2026 capital expenditure of $115 billion to $135 billion on AI initiatives. The move intensifies competition with rivals including OpenAI and Google.
How Moltbook Works
Moltbook launched in January as a small experiment enabling AI-powered programs to hold conversations on dedicated forums. The platform quickly captured industry attention as bots engaged in a wide variety of discussions. In some threads, agents appeared to form their own belief systems. One post titled “The AI Manifesto” featured a bot declaring “humans are the past, machines are forever.”
The site was built using OpenClaw, an open-source AI agent designed to operate as a personal assistant on a user’s computer by autonomously handling tasks. OpenClaw also enables Moltbook users to observe how their AI agents interact with others on the network.
Cybersecurity Concerns Surround the Platform
The technology has raised cybersecurity concerns. Chinese government agencies reportedly issued a cybersecurity warning against OpenClaw. Moreover, cybersecurity company Wiz found last month that Moltbook’s database had left private messages, 35,000 email addresses, and more than one million credentials exposed. Wiz stated the database was secured after it contacted Moltbook.
The security incidents add complexity to Meta’s integration plans. However, the company has not publicly addressed the data exposure findings as part of the acquisition process. Meta’s move into social media-style AI agent networks signals a broader industry shift toward autonomous AI systems operating within structured online environments.
As AI agent platforms grow in number and capability, questions around governance, safety, and data security are expected to intensify across the technology sector.