12,445 papers · continuously updated · last export: 10 Aug 2026livingmeta.ai
← Browse all papers
AI evidence extraction

Generative AI in Academic Publishing

Aakash Kumar, Debra McKeown, Hassan Syed, Cheryl J. Craig, William H. Rupley, Afaq Ahmed et al. · 2026

AI-generated evidence extraction, verified across multiple analytical personas. Not a substitute for the peer-reviewed original.

9/10
Relevance
1/4
Quality (LMQS)
I
Evidence
0
Citations
0.00
FWCI

This is an AI analysis. Read the peer-reviewed original at the publisher: https://doi.org/10.4018/407619

Methodology & findings

Study design

Qualitative comparative analysis of policy documents from five academic publishers (Taylor & Francis, Elsevier, Sage, Wiley, and Springer Nature) using Committee on Publication Ethics (COPE) position statements as an analytical framework..

Sample

N = 5, 1 group

Primary method

Qualitative comparative analysis using thematic coding framework derived from COPE position statements. No quantitative statistical methods reported.

Main result

The study found "substantial differences among publishers in defining acceptable AI use and disclosure requirements." Additionally, "while all publishers explicitly state that AI cannot claim authorship, their policies differ in permitted AI roles, particularly regarding peer review and manuscript preparation."

Reports effect sizes.

Research paradigm

Qualitative-interpretive

Author conclusions

The authors conclude that "the study recommends clearer, discipline-specific guidelines and enhanced reviewer training to ensure responsible AI use for upholding scholarly integrity."

Risk of bias

Selection bias: Only five major publishers analyzed; smaller or regional publishers not included; Document analysis bias: Reliance on published policy statements without validation through author interviews; Temporal bias: Snapshot analysis of policies that may be rapidly evolving; Selection bias: only five major publishers analyzed; smaller publishers and non-English publishers not included; Temporal bias: policies are evolving; snapshot analysis may not capture current state; Interpretation bias: qualitative analysis subject to researcher interpretation of policy documents; No inter-rater reliability reported for thematic analysis

Open questions raised

  • The authors identify a need for "clearer, discipline-specific guidelines and enhanced reviewer training" to address inconsistencies in AI policies across publishers and ensure responsible AI use in academic publishing.
  • The authors identify the need for clearer, discipline-specific guidelines and enhanced reviewer training to address inconsistencies in AI policies across publishers and to ensure responsible AI use in scholarly publishing.
  • Need for clearer, discipline-specific guidelines for AI use in academic publishing; need for enhanced reviewer training on responsible AI use.
Data: not_statedCode: not_statedExtracted from: pdfAgreement 72%

Explore related topics

Institutional PolicyPeer ReviewContent Analysis

Related papers

← Browse all papers