Comprehensive Editorial Protocols for Detecting Ai-Generated Manuscripts in Scientific Journals
Description
The rapid proliferation of Large Language Models (LLMs) such as ChatGPT, Claude, Gemini, and others has introduced an unprecedented challenge to the integrity of scientific publishing. While AI tools can legitimately assist authors in improving language clarity, organizing ideas, or checking grammar, the submission of manuscripts that are substantially or entirely drafted by AI — without meaningful intellectual contribution from the listed authors — violates the fundamental principles of authorship, accountability, and research integrity.
Abstract (English)
The objective of this document is to equip editorial boards, handling editors, and peer reviewers worldwide with a comprehensive, evidence-based, and ethically grounded set of protocols to identify, evaluate, and respond to manuscripts suspected of being predominantly AI-generated with minimal genuine author effort. These guidelines synthesize current best practices from major publishers (Elsevier, Nature Portfolio, Springer Nature, Wiley, Science, JAMA) and ethical frameworks from COPE, ICMJE, and WAME.
Files
MS_Editorial_AI_Detection_Protocols_2026.pdf
Files
(4.3 MB)
| Name | Size | Download all |
|---|---|---|
|
md5:382697109aab4406a1125a3fa9033c28
|
4.3 MB | Preview Download |
Additional details
Dates
- Issued
-
2026-04-07