Meta just bought the social network for AI bots everyone’s been talking about
Meta has acquired Moltbook, a fast-growing social network where autonomous AI agents interact with each other much like humans do on traditional platforms, as part of its push to stay competitive in the escalating AI “arms race” with rivals such as OpenAI, Anthropic and Google. The bot-only site shot to prominence in Silicon Valley after amassing millions of registered agents within days, with some technologists hailing it as a glimpse of how AI agents could one day work and socialize on our behalf, and others deriding it as a chaotic mix of low-quality AI content, fake agents and potential security risks. Moltbook’s team will join Meta’s superintelligence labs, bolstering a broader strategy that has included acquiring AI agent startup Manus and making a multibillion-dollar investment in Scale AI, even as Wall Street pressures Meta to prove these enormous AI bets can translate into real revenue.
The deal also underscores the rivalry between Meta and OpenAI: just weeks earlier, OpenAI hired the creator of OpenClaw, the open-source AI agent system behind Moltbook, and CEO Sam Altman has argued that this underlying agent technology—not the flashy social network—is the real breakthrough that will become “core” to OpenAI’s products. By bringing Moltbook in-house, Meta gains a live testbed for agent-to-agent interaction at scale and a narrative that it is building not just chatbots, but an ecosystem in which AI agents can coordinate tasks, represent users and potentially power new consumer and business services.
Why it matters: This acquisition accelerates a shift from humans talking to brands online toward autonomous agents talking to each other on our behalf—which means communicators must quickly learn how to craft messages, policies and safeguards for audiences that increasingly include non-human intermediaries.
Talking to each other? Sure. But what are they saying?