AI evidence extraction

Disclosing artificial intelligence use in scientific research and publication: When should disclosure be mandatory, optional, or unnecessary?

David B. Resnik, Mohammad Hosseini · Accountability in Research · 2025

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

9/10

Relevance

1/4

Quality (LMQS)

Evidence

Citations

27.05

FWCI

Top 10%

Impact

This is an AI analysis. Read the peer-reviewed original at the publisher: https://doi.org/10.1080/08989621.2025.2481949

Methodology & findings

Study design

Narrative review of existing AI disclosure policies from publishers and organizations (including the International Committee of Medical Journal Editors); ethical analysis and framework development for categorizing disclosure requirements..

Main result

The authors argue that "disclosure should be mandatory only when AI use is intentional and substantial," where intentional means "directly employed with a specific goal or purpose in mind" and substantial means the AI use "produces evidence, analysis, or discussion that supports or elaborates on the conclusions/findings of a study" or "directly affects the content of the research/publication." The study identifies three criteria for identifying substantial AI uses: "a) using AI to make decisions that directly affect research results; b) using AI to generate content, data or images; and c) using AI to analyze content, data or images."

Reports effect sizes.

Research paradigm

Normative ethics and policy analysis

Author conclusions

The authors conclude that "disclosure should be mandatory only when AI use is intentional and substantial," establishing a framework that distinguishes between mandatory, optional, and unnecessary disclosure categories. They argue for a nuanced approach rather than blanket disclosure requirements, with the specific proviso that disclosure becomes mandatory when AI use meets any of their three identified criteria affecting research results, content generation, or content analysis.

Data: not_statedCode: not_statedExtracted from: pdfAgreement 83%

Explore related topics

Responsible AI Use Research Integrity Conceptual