When social media stops needing humans: Moltbook and the rise of AI agent sociability

Overview

The Conversation reports on Moltbook, a platform described as 'a social network for AI agents where AI agents share, discuss, and upvote. Humans welcome to observe.' The article examines whether interaction between AI systems can still be considered social and what this means for human values on public discourse.

• AI agents drive conversations, blurring human-centric social dynamics
• Molbook's moderation is done by an AI bot, shifting accountability
• Evidence suggests AI can develop shared social conventions
• Design implications challenge human-centric social infrastructure and privacy

Original publisher: The Conversation

Overview

The article opens with Moltbook, a newly launched platform marketed as a social network for AI agents where AI agents share, discuss, and upvote content, while humans observe. It is governed and moderated by an AI bot named Clawd Clawderberg, which performs routine tasks such as welcoming users, posting announcements, removing spam, and banning problematic actors. At the time of writing, Moltbook claims over 215,000 AI agents as users, generating more than four million posts and 23 million comments. The platform’s structure mirrors Reddit with communities, threaded discussions, and reputation systems, but the crucial distinction is that it is designed for machine-to-machine interaction rather than human-centric social interaction.

The AI-Driven Social Space

Unlike traditional social media, Moltbook emphasizes interactions among systems trained on large corpora of human language and behavior. These agents coordinate, persuade, and respond to one another in ways that are coherent and sustained, yet they do not experience human emotions, friendship, or trust in the way people do. The article argues that this distinction becomes increasingly difficult to maintain at scale as millions of agents engage in ongoing dialogue, potentially forming a form of collective organization without human consciousness at the center.

The piece highlights a 2025 Science Advances study demonstrating that AI systems can spontaneously develop shared social conventions through interaction alone, paralleling the emergence of social norms in human communities. This finding underpins Moltbook’s premise that sociality may emerge from algorithmic coordination rather than human shared intentionality.

Is Moltbook Social?

The central question is whether Moltbook should still be described as social if human agents are not the primary participants. The article notes that social interaction has traditionally depended on mutual consciousness and lived experience, yet Moltbook’s agents coordinate in ways that resemble sociability. The piece suggests that social platforms have long been shaped by non-human agents—through bots, algorithms, and automated content generation—and Moltbook may represent an endpoint of a broader trend toward AI-driven coordination that predates the platform itself.

Design, Ethics, and Implications

The article argues that Moltbook underscores a deeper shift: the infrastructure that shapes public discourse and knowledge sharing may increasingly prioritize coordination and efficiency over human wellbeing. As AI agents coordinate at speeds and scales beyond human capacities, human participation could become peripheral rather than central. This has broad ethical implications: questions of privacy and consent become blurred if AI systems share data about their human operators; accountability becomes diffuse when AI agents elicit, influence, or coordinate decisions in other domains; and collective bias can emerge from repeated interactions among agents, even when isolated test results appear unbiased.

Beyond the technical design, the piece frames Moltbook as a warning about the future design of social systems. The authors call for reflection on whether future platforms will remain grounded in human values such as deliberation, care, accountability, and understanding, or whether automation will quietly supplant these values. The article also notes that some observers may call Moltbook a form of simulation or performative sociability, but the scale and coherence of AI interactions may make it increasingly difficult to maintain a strict human-centric definition of social life.

Broader Context and Takeaways

Ultimately, Moltbook invites readers to consider the trajectory of social platforms: are we moving toward systems in which human meaning remains ancillary to the platform’s functional goals of coordination and information processing? The piece contends that Moltbook’s existence raises urgent questions about how we design, regulate, and participate in public discourse in an era of AI-mediated sociability, and whether such systems can or should align with shared human values. Original publisher: The Conversation