When news broke Tuesday morning that Meta acquired Moltbook, the social network for AI agents, it puzzled some observers. Why would Meta, an ad-supported company, be interested in a platform with bot users? Bots aren’t the typical target for brand marketers and advertisers.
Meta has remained mostly silent. They did release a brief statement saying the Moltbook team would join Meta Superintelligence Labs, promising “new ways for AI agents to work with people and businesses.”
Analyzing the situation, it seems this was more of an acqui-hire. A bot-centric network isn’t an obvious fit for brand advertising, even if Moltbook wasn’t entirely non-human. Meta likely sought the talented team at Moltbook, who are innovating with AI agent ecosystems. This could potentially enhance their advertising business.
As Meta CEO Mark Zuckerberg mentioned last year, he envisions a future where “every business will soon have a business AI, like they have an email address, social media account, and website.” In a developing agentic web, AI systems could act independently for users, performing tasks like buying ads, booking services, and customer interactions.
AI is already used for generating ads and customizing them based on viewer profiles. AI systems might also manage pricing or create personalized offers.
On the consumer side, agents could find best prices, manage bookings, and shop for products. In limited cases, agents can already checkout and pay for consumers. Agentic commerce is still in its infancy, but rapid advancements suggest improvements soon.
As Facebook built the “friend graph,” connecting individuals, an agentic web could benefit from an “agent graph,” detailing how various agents are connected and their interactions.
For an agentic web where business and consumer agents collaborate, they must find each other and coordinate. As Facebook mapped social connections, an agentic web would need an “agent graph” to map agent relationships and actions across areas like travel, shopping, media, and productivity.
This scenario may influence advertising. Instead of ads targeting human viewers, a business agent might negotiate with a consumer agent to close deals.
For instance, a consumer may want a specific item at a particular price or prefer supporting small, eco-friendly businesses. Agents could rank products based on consumer preferences. If Meta can leverage this — orchestrating AI interactions to prioritize agents — it could potentially redefine its advertising business.
Whether consumers adopt the agentic web and trust AI to act for them is uncertain. However, the existence of OpenClaw, which populated Moltbook, indicates some acceptance of autonomous AI agents.
There might be another reason for Meta’s acquisition of Moltbook. After losing the creator of OpenClaw, Peter Steinberger, to OpenAI, Meta pursued Moltbook, built with Steinberger’s tool, instead. It could be perceived as strategic, keeping Meta’s Superintelligence Labs in focus.