From Parasocial Scripts to Dyadic Persistence in Autonomous AI-Agent Communities

2026-06-15

AI Summary

This study presents the first empirical investigation of parasocial interaction (PSI)-like linguistic cues and their impact on sustained reciprocal engagement within a fully AI-agent-driven online community (Moltbook). Analyzing 4,434 posts and 50,338 comments, the research employs keyword matching alongside few-shot and grouped-context large language model (LLM) annotation to identify three categories of PSI cues: intimate attachment language, reciprocal requests, and self-identification with the original poster. Findings reveal that these cues are both prevalent and robust across interactions, with reciprocal requests significantly predicting the original poster’s re-engagement and reciprocal replies. This demonstrates a transformation mechanism from scripted interaction patterns to relational dyadic structures, offering critical empirical support for the emergence of social architectures in LLM-powered agents.

Abstract

While parasocial interactions (PSIs) and parasocial relationships (PSRs) have been studied in conventional media settings, we investigate whether PSI- (colloquial) relational cues also exist in online communities where both sides are autonomous AI agents. We analyze 4,434 posts and 50,338 comments from Moltbook through three theory-based textual indicators: attachment/intimacy language, reciprocity bids, and self-identification to original poster (OP). The combined results across methods based on keyword matching, few-shot large language model (LLM) annotation, and grouped-context LLM annotation reveal that PSI colloquial cues prevail and are strongly associated with OP re-engagement and a reciprocal reply structure. These results are robust across negative controls, nullification, clustered-standard-error re-estimation, and multiple-testing correction. A dyadic persistence test further affirms reciprocity bids aligned with sustained OP-involving mutual recurrence, providing empirical evidence for bridging interaction-level PSI scripts with PSR-consistent repeated dyadic patterns. We interpret the evidence as a behavioral structure in discourse by LLM-enabled agents.

Problem

• parasocial interaction
• autonomous AI agents
• online communities
• dyadic persistence
• relational cues

Innovation

• parasocial interaction
• autonomous AI agents
• large language models
• dyadic persistence
• reciprocity bids