Student or educator? Sign up with your school email and get 6 months of Pro free.

The DOI matches its Crossref record; 38 of the 38 references it cites were found in a registry or an index; 18 of the 29 sources its claims cite were checked for support.

The Attention Mismatch: Mapping the Structural Academic Governance Deficit in the Age of Generative AI

Open access
Guo, Zhenning, Mao, Haoran, Zhang, Fang · 2026 · Publications

Abstract

With the rapid advancement in Generative Artificial Intelligence (GenAI), AI-generated content (AIGC) lacking human cognitive oversight is increasingly permeating open web environments and academic communication systems. This study integrates longitudinal retraction data (Retraction Watch Database, 1990–2026), web-scale analyses of AI-content penetration (Common Crawl, 2013–2026), and bibliometric mapping of governance scholarship (Web of Science Core Collection, Scopus, Google Scholar, 2020–2026) to diagnose the cross-level misalignment between synthetic-content diffusion, AI-related misconduct pressure, and governance attention. On this basis, it proposes a Normalized Coverage Index (NCI) to measure the relative relationship between scholarly attention to AI-related academic misconduct governance and the level of misconduct pressure observed through retraction data across disciplines. The results reveal pronounced asymmetries at the disciplinary level. Fields such as chemistry (0.04), physics, mathematics & statistics (0.11), and life sciences & biology (0.34) exhibit clear governance gaps, whereas Education shows a comparatively excessive level of attention (NCI = 29.26). Since 2022, AIGC has expanded rapidly across open web corpora, accompanied by a sharp rise in AI-related retractions, which also exhibit a longer detection lag than traditional forms of misconduct (2.77 years vs. 1.91 years). Although the volume of academic governance-related research has grown rapidly, its proportion within the broader body of AI-related research has declined, suggesting that scholarly attention to governance has not kept pace with technological diffusion. Consequently, a structural misalignment in governance—closely tied to the allocation of attention—has emerged within the academic system in the era of GenAI. This misalignment may pose potential risks to the robustness of the knowledge production system. Addressing it requires rebuilding epistemic infrastructure through provenance transparency, auditable workflows, and governance-aware seed corpora aligned with empirically concentrated risks.

Overview

CoChat Summary

TL;DR

Integrating retraction records, web-corpus AI-detection, and bibliometric mapping, this study argues that governance research on AI-related academic misconduct is structurally misaligned with observed misconduct pressure across disciplines. It proposes a Normalized Coverage Index (NCI) showing that basic sciences are under-covered (chemistry NCI = 0.04) while education is heavily over-covered (NCI = 29.26).

Key findings

  • AI-related retractions surged after 2022, and their detection lag exceeds that of traditional misconduct: mean 2.77 years vs. 1.91 years, median 1.62 vs. 0.24 years (Mann-Whitney U and KS tests p < 0.001; Cohen's d = 0.25). Subcategory delays were longest for Data & Result Fabrication (mean = 4.472 years; median = 2.916) and Plagiarism & Duplicate Publication (mean = 3.390), shortest for Article Quality & Citation Issues (mean = 1.779).
  • Average AI-likeness of Common Crawl web text stayed low and stable during 2013–2021, then rose markedly from 2022, persisting at 0.2660–0.3915 through 2026. The ensemble detector achieved accuracy 0.9947, F1 0.9969, ROC-AUC 0.9995, with a 0.53% test-set misclassification rate.
  • NCI values: Chemistry 0.04 (95% CI 0.00–0.08), Physics, Mathematics & Statistics 0.11 (0.04–0.22), Economics, Business & Management 0.31, Environmental, Earth & Agricultural Sciences 0.33, Life Sciences & Biology 0.34, Engineering & Materials 0.42, Medicine & Health Sciences 0.57 — all below 1. Above 1: Computer Science & IT 2.08, Law, Ethics & Policy 6.03, Education 29.26 (95% CI 26.68–32.14).
  • Governance publications grew rapidly but declined as a share of overall AI research. Output is dispersed (top 15 venues = 19%, 500/2566) and geographically concentrated (top 15 countries = 59%, 1508/2566). Top keywords: "Artificial Intelligence" 716 (15.65%), "ChatGPT" 520 (11.36%), "Generative AI" 478 (10.44%), "Academic Integrity" 439 (9.59%), "Higher Education" 333 (7.28%); top five = 54.32%.

Methods

  • Retraction analysis used the Retraction Watch Database (1990–2026), classifying AI-related retractions into five categories (AI Content Generation & Abuse; Data & Result Fabrication; Plagiarism & Duplicate Publication; Authorship & Process Violations; Article Quality & Citation Issues), with retraction lag compared via Gaussian KDE (bandwidth = 0.3), Mann-Whitney U, KS tests, and Cohen's d.
  • AIGC penetration was measured on Common Crawl web text (2013–2026, extracted 30 January 2026, fixed seed = 42), training an ensemble of word-level (1–2 g) and character-level (3–5 g) TF-IDF logistic regression classifiers weighted 0.6/0.4, with human negatives from 2013–2021 English web text and AI positives from HC3; five-fold cross-validation, implemented in Python 3.11.
  • Bibliometric mapping drew on WoSCC, Scopus, and Google Scholar (1 January 2020–30 January 2026), with keyword co-occurrence in VOSviewer 1.6.20 and Kleinberg burst detection (s = 2.0, γ = 1.0). NCI compared governance publication share to retraction share across 13 harmonized disciplines (Cohen's κ = 0.81; ε = 0.01; 2000-iteration bootstrap; Spearman's ϱ > 0.9 for robustness checks).

Limitations

  • Stated: the AI-likeness model is trained on existing corpora and may be affected by model evolution, so results indicate trends rather than precise proportions of generated content; the Common Crawl sample is limited relative to the full corpus and web-content heterogeneity may limit generalizability; retraction data as a misconduct proxy are biased by disciplinary differences in detection, reporting, and time windows, creating "detection-dependency circularity" and possible underestimation of governance needs in low-NCI fields.
  • Stated: the study includes only retracted publications, excluding cases under investigation or undetected, so retraction lag is a conservative estimate; NCI captures attention distribution, not governance effectiveness.
  • Evident from the text: country-level retraction counts are absolute and not normalized by publication volume, so the authors caution they should not be read as relative research-integrity levels; the link between AI-like content and misconduct is described as associational and requiring causal validation.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.3390/publications14020027

Cites

38 of 38 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

14 Valid · 6 Not valid · 8 Not checked · 133 Own results and interpretations

81 more claims · 5 with their cited sources checked

Create a free account to see all 161 claims and every source verdict.

See all claims — free account
  • It may state two separate claims as one. 59
  • Probably not a substantial claim. 25
  • The quote only partly states this claim. 1

Create a free account to see these claims.

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Self-check results appear here once an independent check has confirmed them.

Cited works that were retracted

Works this paper cites that were later retracted, from the retraction notices publishers deposit. Major when the notice came out before this paper was published.

1 of 1

Citations

Cites

1 retracted

Show all 35 references

Metadata

Authors
Guo, Z., Mao, H., & Zhang, F.
Affiliations
China University of Petroleum, East China (CN)
Corresponding author
Mao, Haoran
Venue
Publications
Publisher
MDPI AG
Published
2026-04-17
Volume
14
Issue
2
Pages
27
ISSN
2304-6775 (ISSN-L)
OpenAlex
W7154689640
Cited by
0
References
35
FWCI
0.00
Type
journal article
Language
en
Open access
Locations
  • Publications (publishedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
Funders
General Project of the National Social Science Fund of China (25BGJ095)
Topics
  • Social Sciences › Education › Academic integrity and plagiarism
  • Social Sciences › Library and Information Sciences › Academic Publishing and Open Access
  • Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
Keywords
generative AI, AI-generated content, academic misconduct, research integrity, university governance, retraction, disciplinary differences, knowledge production

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-06, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-06 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: The Attention Mismatch: Mapping the Structural Academic Governance Deficit in the Age of Generative AI. https://app.cochat.ai/papers/10.3390/publications14020027
The paper
Guo, Z., Mao, H., & Zhang, F. (2026). The Attention Mismatch: Mapping the Structural Academic Governance Deficit in the Age of Generative AI. Publications. https://doi.org/10.3390/publications14020027

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.