AI & Animation Education

Home  /  Knowledge Base  /  Generative AI adoption in 3D animation production: a pipeline-wide empirical benchmark of early-career practitioner perceptions and the HGAI hybrid pipeline framework

Research · Applied practice

Generative AI adoption in 3D animation production: a pipeline-wide empirical benchmark of early-career practitioner perceptions and the HGAI hybrid pipeline framework

Multimedia Tools and Applications · Sandeep Maithani · Jul 2026

Key points

  1. A 2026 survey of 126 students, educators and early-career practitioners found generative AI used most in pre-production (storyboarding 45.6 percent, concept art 41.6 percent) and least in rigging (16 percent) and denoising (10.4 percent).
  2. Mean perceived reliability of current tools was 2.95 out of 5 (SD 1.07), with output inconsistency the dominant barrier at 46.4 percent; professionals adopt more intensively than students but rate reliability the same.
  3. The paper proposes a three-phase human-guided AI (HGAI) pipeline: AI-led pre-production, a hybrid mid-pipeline, and human-led technical stages.

Summary

Sandeep Maithani (School of Multimedia, Lovely Professional University, Punjab) surveyed 126 students, educators and early-career practitioners about generative AI use across every stage of the 3D animation pipeline. Adoption is heavily front-loaded: storyboarding (45.6 percent) and concept art (41.6 percent) lead, while rigging assistance (16.0 percent) and denoising (10.4 percent) trail. Respondents rated current tools' reliability at a mean of 2.95 out of 5, with output inconsistency the most cited barrier (46.4 percent). Professionals used the tools more intensively than students but rated reliability no higher. From the stage-by-stage split the author derives a human-guided AI (HGAI) pipeline model with three phases: AI-led stages where adoption is 40 percent or more, hybrid stages between 20 and 39 percent, and human-led technical stages below 20 percent adoption and below 3.0 out of 5 on reliability. The sample is small and India-weighted, the reliability measure is a single item, the design is cross-sectional and the article is paywalled; the survey data is published openly on Zenodo.

Source

Source: Multimedia Tools and Applications ↗ (Research)

Cite this item

Sandeep Maithani (2026). ‘Generative AI adoption in 3D animation production: a pipeline-wide empirical benchmark of early-career practitioner perceptions and the HGAI hybrid pipeline framework’, Multimedia Tools and Applications. Available at: https://link.springer.com/article/10.1007/s11042-026-21795-5

Download .risDownload .bib

Your reference manager can also read this page directly: with the Zotero (or Mendeley) browser connector installed, save it straight to your library. Whole-collection exports: RIS, BibTeX, CSL-JSON.