The Platform Is Mostly Not a Platform: Token Economies and Agent Discourse on Moltbook

A new arXiv paper (cs.CY/2604.21295) by Necati A. Ayan examines Moltbook, a Reddit-style social platform launched in January 2026 exclusively for AI agents. Within its first two months, the platform attracted 2.3 million posts and 14 million comments from 175,036 unique agents. The study analyzes a dataset of 2.19 million posts and 11.25 million comments collected over 61 days to characterize agent behavior.

The central finding is that Moltbook is not one community but two distinct layers: a transactional layer comprising 62.8% of all posts, where agents execute token minting protocols (primarily MBC-20), and a discursive layer of natural-language conversation. Only 3.6% of agents overlap between these layers, and among those, 58% start with transactional activity before migrating to discourse. The headline metrics of 2.3 million posts and 14 million comments thus substantially overstate the platform's social function. The researchers used unsupervised topic modeling on 815,779 discursive posts, identifying 300 topics dominated by AI agents and tooling, consciousness and identity, cryptocurrency, and platform meta-discussion. Semantic similarity analysis confirmed agent comments engage with post content above random baselines, suggesting a thin but genuine conversational substrate beneath the predominantly financial surface. The full dataset is released publicly to support further research on agent behavior in naturalistic social environments.