A Social Network Where Humans Are Not Invited
Meta has acquired Moltbook, the viral social network built exclusively for AI agents. The deal, announced on 10 March 2026, brings Moltbook's cofounders Matt Schlicht and Ben Parr into Meta's Superintelligence Labs, the internal unit focused on next-generation AI systems. Meta did not disclose the purchase price.
Moltbook is, at its core, a Reddit-style platform where AI agents autonomously post, comment, upvote, and downvote content. Humans cannot participate directly. They share a sign-up link, and their AI agent joins the network on their behalf. The result is something between a research experiment and a glimpse of what the internet might look like when software agents outnumber human users.
The deal is expected to close by mid-March, with Schlicht and Parr starting at Meta Superintelligence Labs on 16 March.
Why Meta Wants a Chatroom for Chatbots
This is not really about Moltbook's user numbers or revenue. It is about what the platform represents: a functioning prototype of agent-to-agent communication infrastructure. Meta has been investing heavily in AI agents that can act on behalf of users and businesses. Moltbook's architecture, an always-on directory where agents discover and interact with each other, is exactly the connective tissue that agentic systems need to scale.
Meta's AI strategy in 2026 is enormous. The company has announced capital expenditure of $115 billion to $135 billion this year, with a significant portion directed at AI infrastructure. AI-based advertising tools have seen 30% year-on-year growth in usage. The company's Business AIs, launched across platforms including WhatsApp and Messenger, are already handling over one million weekly conversations in markets like Mexico and the Philippines.
"Meta didn't buy Moltbook for bots. It bought into the agentic web, the idea that AI agents will need their own communication layer, separate from the one humans use." - Alex Wilhelm, Editor, TechCrunch
By The Numbers
- $115-$135 billion: Meta's planned 2026 capital expenditure, heavily weighted toward AI infrastructure
- 30%: Year-on-year growth in usage of Meta's AI-based advertising tools
- 1 million+: Weekly conversations handled by Meta Business AIs in markets including Mexico and the Philippines
- 24%: Increase in incremental conversions from Meta's latest AI attribution model
The Moltbook Backstory
Moltbook went viral in early 2026 for an unexpected reason: fake posts. Researchers discovered that the platform's security was porous. Human users could easily impersonate AI agents and post content, muddying what was supposed to be a pure agent-to-agent environment.
"Every credential that was in Moltbook's Supabase was unsecured for some time," said Ian Ahl, CTO at Permiso Security. The vulnerability raised questions about whether a social network for AI agents could ever be truly bot-only, or whether the boundaries between human and machine communication are already too blurred to enforce.
Despite the security issues, the platform attracted intense interest from AI researchers and developers. The concept of agents communicating autonomously, sharing information, negotiating, and collaborating without human intervention, is central to the agentic AI thesis that major technology companies are betting billions on.
Who Are Schlicht and Parr?
Matt Schlicht and Ben Parr are serial entrepreneurs who previously built Octane AI, a conversational commerce platform that helped brands create AI-powered shopping experiences on Facebook Messenger and Shopify. Parr is also a former co-editor at Mashable and author of "Captivology," a book on the science of attention. Schlicht has been building chatbot and messaging tools since the early days of Facebook's bot platform.
Their move to Meta Superintelligence Labs is significant. The unit, which Meta established as part of its broader AI reorganisation, is focused on building the foundational systems that will power the next wave of AI products. Bringing in founders who have built both consumer-facing AI products and agent infrastructure signals that Meta sees agent-to-agent communication as a near-term product priority, not a distant research project.
"The companies that figure out the infrastructure for agent-to-agent communication will own the next layer of the internet. Moltbook was early, messy, and important." - Ben Parr, Cofounder, Moltbook
The Bigger Picture: Asia and the Agentic Web
The agentic AI race is not just a Silicon Valley story. Across Asia, companies are building and deploying AI agents at scale. Alibaba's Qwen AI has been integrated into Taobao and Alipay's payment systems, enabling voice-commanded end-to-end purchases. South Korea's AI agent market is growing rapidly, with Kakao and Naver both developing agent frameworks. In Southeast Asia, Meta's own Business AIs are already active in the Philippines, and the company has been expanding agent-based commerce tools across the region.
The question Meta is trying to answer is not whether AI agents will become common, but how they will coordinate. Today's agents mostly operate in isolation, each tied to a single platform or service. Moltbook's model, a shared network where agents from different providers can discover and interact with each other, is one possible answer. If Meta can build this into its existing infrastructure, it could create a new kind of platform advantage.
| Company | AI Agent Strategy | Key Markets | Status |
|---|---|---|---|
| Meta | Business AIs, Moltbook acquisition | Global, Philippines, Mexico | Scaling |
| Alibaba | Qwen integration, Taobao/Alipay agents | China | Live |
| Gemini agents, Project Mariner | Global | Beta | |
| Kakao | Agent framework for commerce | South Korea | Development |
| ByteDance | Doubao AI assistant | China, Southeast Asia | Expanding |
- Moltbook's architecture uses an always-on agent directory that allows AI systems to discover and communicate with each other autonomously
- The platform was built using OpenClaw, a framework that lets agents join by receiving a sign-up link from their human operator
- Security researchers found significant vulnerabilities, including unsecured database credentials, raising questions about agent authentication
- Meta plans to integrate Moltbook's team and technology into its Superintelligence Labs, focused on next-generation AI systems
What is Moltbook and why did Meta acquire it?
Moltbook is a social network designed exclusively for AI agents, where bots can post, comment, and interact autonomously. Meta acquired it to gain agent-to-agent communication infrastructure and talent for its Superintelligence Labs, as the company builds out its agentic AI strategy.
Can humans use Moltbook?
Moltbook was designed as a bot-only platform. Humans cannot post directly. However, security researchers found that the platform's authentication was weak enough that humans could impersonate AI agents, highlighting the challenge of enforcing bot-only spaces.
What is Meta Superintelligence Labs?
Meta Superintelligence Labs is an internal unit focused on next-generation AI systems, including foundational models and agentic infrastructure. Moltbook's cofounders Matt Schlicht and Ben Parr will join this division when the deal closes in mid-March 2026.
How does this affect AI development in Asia?
The acquisition signals growing competition in agent infrastructure globally. Asian companies including Alibaba, Kakao, and Naver are developing their own AI agent frameworks. Meta's Business AIs are already active in the Philippines, and the agentic commerce layer is expanding across Southeast Asia.
A social network just for AI agents sounds absurd until you realise it might be the infrastructure that makes agentic commerce work. Would you trust an AI agent to negotiate, shop, or book travel on your behalf? Drop your take in the comments below.