Generative artificial intelligence addiction is prospectively associated with psychotic-like experiences: Evidence from a two-wave longitudinal mediation and network study

Key Findings

• — Baseline generative-AI addiction was associated with psychotic-like experiences at one-month follow-up (beta = 0.19, p < .001).
• — Rumination was a significant mediator; information overload was not an independent mediator but contributed through a serial path via rumination.
• — Cross-lagged panel network analysis showed baseline psychotic-like experiences relatively strongly associated with rumination-related nodes at the second wave.
• — The authors present this as longitudinal evidence that cognitive processes, particularly rumination, may partly link problematic generative-AI use to psychotic-like experiences.

Methodology Notes

Two-wave longitudinal survey, n = 582 adolescents and young adults (China), one-month interval, self-report scales for GAI addiction, information overload, rumination and psychotic-like experiences; serial mediation models and cross-lagged panel network analysis. Online publication date 2026-09-09 per OpenAlex; Crossref assigns the December 2026 issue. Open access. ScienceDirect blocks automated fetching from this workspace, so the record was built from the OpenAlex deposited abstract and Crossref metadata; full text not read. Short follow-up interval and a single-country student-age sample limit inference; the measure is psychotic-like experiences, not clinical psychosis.

Sources