AI evidence extraction

RESPONSIBLE USE OF AI-GENERATED CONTENT IN VIETNAMESE SCHOLARLY PUBLISHING: EVIDENCE FROM JOURNAL POLICIES AND EDITORIAL PRACTICES

Tran Huu Tuyen · Veredas do Direito Direito Ambiental e Desenvolvimento Sustentável · 2026

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

9/10

Relevance

3/4

Quality (LMQS)

Evidence

Citations

0.00

FWCI

This is an AI analysis. Read the peer-reviewed original at the publisher: https://doi.org/10.18623/rvd.v23.n1.4029

Methodology & findings

Study design

Exploratory mixed document-analysis approach combining structured policy clause review of five Vietnam-affiliated publishing contexts (university-based journals, defense-related journal, law/social-science journal, internationally co-published journal, and multidisciplinary open access journal) and limited practice evidence from published AI declaration statements.

Sample

N = 5, 6 groups

Primary method

Structured qualitative content analysis using two-pass coding procedure. First pass identified policy clauses; second pass enabled cross-case comparison and thematic synthesis. Coding scheme dimensions captured operational questions and example indicators for each policy domain.

Main result

The study found that "across cases, we observe partial convergence on core principles: human accountability remains central and AI tools are not granted authorship status" and identified "substantial fragmentation in disclosure requirements, treatment of AI-generated images and references, restrictions on reviewer use of AI tools, and clarity of enforcement mechanisms" across Vietnam-affiliated journals. Additionally, "Vietnam-affiliated governance is best characterized as 'conservative convergence with operational unevenness,' [where] journals tend to adopt protective norms, but implementation clarity varies—especially for peer review and enforcement."

Reports effect sizes.

Research paradigm

Qualitative-interpretive with policy-focused governance analysis

Author conclusions

The authors conclude: "This paper examined responsible AIGC governance in Vietnam-affiliated scholarly publishing through an exploratory mapping of journal policies and observed editorial practices. The evidence indicates partial convergence around international baseline norms—particularly the rejection of AI authorship and the emphasis on human accountability—yet significant fragmentation in disclosure rules, peer review safeguards, and enforcement clarity." They propose that "rather than recommending blanket bans, we argue for proportional governance: low-risk uses (language correction, translation) can reduce inequality, while high-risk uses (peer review drafting, data and image fabrication) require strict prohibition or rigorous methodological justification."

Risk of bias

Selection bias: purposive sampling of 5 cases may not represent full diversity of Vietnamese journals; Availability bias: study limited to publicly accessible policy documents, missing internal editorial guidelines; Temporal bias: policies change rapidly; snapshot captures only current state; Document source bias: reliance on published policies may overrepresent formally structured journals; Language bias: potential undersampling of smaller domestic journals not in English or with limited online presence; Selection bias: only five Vietnam-affiliated journals sampled (non-representative sampling); Document bias: analysis relies on publicly available policy documents that may not reflect actual implementation practices; Temporal bias: policies are rapidly evolving; snapshot is time-bound; Language bias: documentation primarily in English may not capture all Vietnamese-language journals; Publication bias: journals with published AI declaration statements overrepresented in practice sample; Selection bias: Purposive sampling of five Vietnam-affiliated journals may not represent full diversity of Vietnamese publishing ecosystem; Documentation bias: Reliance on publicly accessible policy documents may miss unpublished or informal governance practices; Temporal bias: Snapshot at specific time point; policies in rapid flux; Language bias: Analysis of English-language policy text may not capture nuances in Vietnamese-language journals

Limitations

This study has multiple limitations: "This is an exploratory study relying on publicly accessible documents
Policies change quickly, and journals may implement internal rules not reflected online
Findings should therefore be read as a governance snapshot and a basis for scalable future research rather than a definitive national audit." Additionally, the authors note that "longitudinal monitoring is needed to track how Vietnamese journals update guidance as international norms shift
Second, broader sampling across disciplines and platforms would enable stronger generalization
Third, interviews with editors and reviewers could clarify how policies are implemented in practice, including informal use of AI tools."

Open questions raised

Longitudinal monitoring needed to track how Vietnamese journals update guidance as international norms shift
Broader sampling across disciplines and platforms required for stronger generalization
Interviews with editors and reviewers needed to clarify implementation of policies in practice and informal AI tool use
Large-scale policy audits combined with ethnographic and survey methods to evaluate governance feasibility, author compliance, and reviewer behavior in real editorial systems
Broader sampling across disciplines and platforms would enable stronger generalization
Interviews with editors and reviewers could clarify how policies are implemented in practice, including informal use of AI tools

Data: All datasets relevant to this study's findings are fully available within the article.; All datasets relevant to this study's findings are fully available within the article (as stated by authors)Extracted from: pdfAgreement 65%

Explore related topics

Institutional Policy AI Governance Qualitative