Endocrinology journals lack unified AI standards
A cross-sectional analysis of the top 100 endocrinology journals ranked by the 2023 SCImago Journal Ranking database found that most publications require disclosure of artificial intelligence use and permit AI-assisted manuscript writing, yet few endorse standardized reporting frameworks for AI-driven research.
The findings, which were led by Eva Chen, of the Oklahoma State University Center for Health Sciences in Tulsa, Oklahoma, and colleagues, were published in Frontiers in Endocrinology. Researchers reviewed artificial intelligence (AI) usage guidelines from publicly available Instructions for Authors pages between November 2024 and July 2025, including authorship, manuscript writing, content and image generation, and disclosure requirements.
Most of the journals analyzed were published in Europe or North America, and 61 were ranked in the first quartile by SCImago. Springer Nature, Elsevier, and Wiley Blackwell accounted for 59 journals.
Among the top 100 journals, 84 mentioned AI in their author guidelines, and 79 required disclosure of AI use during manuscript submission. Although no journals permitted AI tools to qualify for authorship, 64 allowed AI-assisted manuscript writing, 22 permitted AI-generated content, and 50 allowed AI-generated images.
Forty journals did not specify policies regarding AI-produced content, and 45 did not address AI-generated images.
Among the 84 journals that mentioned AI in their author guidelines, 94% required disclosure of AI use, 76% allowed AI-assisted manuscript writing, and 45% prohibited the use of AI tools for content generation.
The researchers found no statistically significant association between AI-related policies and SCImago Journal Rank scores.
Endorsement of wider editorial guidance was limited. Nine journals referenced the International Committee of Medical Journal Editors (ICMJE) guidance related to AI, and 12 cited the Committee on Publication Ethics (COPE) recommendations. No journals referenced the World Association of Medical Editors (WAME) recommendations.
Only one journal explicitly required adherence to Consolidated Standards of Reporting Trials–Artificial Intelligence (CONSORT-AI) and Standard Protocol Items: Recommendations for Interventional Trials–Artificial Intelligence (SPIRIT-AI) reporting standards.
The researchers noted that variability across publishers may contribute to inconsistent AI policies, citing the different guidelines for AI-generated images and manuscript assistance, as well as the growing number of available AI reporting frameworks.
Safeguards Against Confusion, Discrepancy
“Our findings, along with current literature, highlight the necessity of standardizing the use of generative AI policies and AI reporting guidelines across all journals,” the researchers wrote. “Journals at a minimum should provide specific statements regarding permissible AI uses to prevent any confusion or discrepancies in future publications.”
Study limitations included possible human error during data extraction, nonstandardized Instructions for Authors pages across journals, and the possibility that some externally linked AI policies were missed. The investigators also acknowledged that their search strategy used a limited range of AI-related terminology.
The researchers reported no conflicts of interest and stated that generative AI was not used in preparing the manuscript.
AACE Endocrine AI is published by Conexiant under a license arrangement with the American Association of Clinical Endocrinology, Inc. (AACE®). The ideas and opinions expressed in AACE Endocrine AI do not necessarily reflect those of Conexiant or AACE. For more information, see Policies.