Truace
Subscribe

Perceived AIGC saturation and AI content fatigue on short-video platforms: An appraisal-based structural equation modeling study

Artificial intelligence-generated content (AIGC) is increasingly embedded in short-video platforms, where algorithmic recommendation systems expose users to high volumes of rapidly produced synthetic media. Although prior research has examined generative AI from the perspectives of creativity, productivity, and technology adoption, less is known about how ordinary users evaluate repeated exposure to AI-generated short videos. This study develops and tests an appraisal-based model of AI content fatigue, distingui…

By the Truvace desk2 min read1 sourcePeer-reviewedTRV-2026-1732· cite & verify
Perceived AIGC saturation and AI content fatigue on short-video platforms: An appraisal-based structural equation modeling study
Wikidata Workshop - Practical part, UM professors - Maastricht University - 15 October 2024 by Olaf Janssen. CC BY 4.0 · https://creativecommons.org/licenses/by/4.0

In brief

A May 2026 cross-sectional online survey of Chinese short-video users tested how people react to repeated exposure to AI-generated videos. After invalid responses were removed, 513 questionnaires were analyzed with confirmatory factor analysis and structural equation modeling, and the adjusted model fit the data acceptably.

The results suggest that fatigue is not driven simply by content being made by AI, but by how users judge its quality and the mental burden it creates. That bears on how platforms handle labeling, recommendation diversity, and moderation, though the cross-sectional design and a single national user group leave open how far the pattern generalizes or whether it holds over time.

Main points

  1. The study analyzed 513 valid questionnaires from Chinese short-video users collected in May 2026.
  2. Perceived AIGC saturation was positively associated with perceived low-quality prevalence, narrative homogenization, and content overload.
  3. Perceived low-quality prevalence had the largest estimated coefficient, followed by content overload.
  4. Narrative homogenization was not supported under bootstrap inference, and its indirect pathways were not supported either.
  5. The authors point to implications for platform governance, AIGC labeling, recommendation diversity, and quality-sensitive moderation.

The problem

Among 513 surveyed Chinese short-video users, perceiving AI-generated content as saturating feeds was tied to fatigue, which was linked to lower satisfaction and lower intention to keep watching.

The rundown

The authors frame AI content fatigue as a content-specific evaluative state. They distinguish perceived AIGC saturation from three user appraisals: low-quality prevalence, narrative homogenization, and content overload.

Model fit statistics were reported as acceptable: chi-square(1321) = 1991.541, chi-square/df = 1.508, CFI = .946, TLI = .929, IFI = .947, and RMSEA = .031.

Bootstrap analyses supported indirect associations from saturation to fatigue, satisfaction, and continued viewing intention through low-quality prevalence and content overload. Those through narrative homogenization were not supported.

Sources

  1. Peer-reviewedPLOS One2026-10-16

The debate