Ethics Policy Society
The AI Guidelines of Scientific Journals Do Not Halt the Use of AI – A Significant Gap in Transparency
A new analysis shows that the guidelines of scientific journals regarding the use of AI do not significantly curb the utilization of AI in academic writing. Text produced or refined by AI is rapidly becoming more common across various scientific fields, regardless of whether the journal has an official AI policy or not.
The study reviewed 5,114 scientific journals and over 5.2 million articles published in them. The majority of the journals – about 70 percent – had adopted some form of AI-related policy. Most often, the guidance was limited to asking authors to disclose if they have used generative AI, such as large language models, for producing or editing text.
However, the data shows that the use of AI is sharply increasing everywhere, and the difference between journals with AI policies and those without is not statistically significant. The growth is particularly rapid in countries where English is not the primary language, as well as in the natural sciences and journals where a large portion of the articles are openly accessible.
Researchers also conducted an in-depth text analysis of 164,000 scientific articles to assess the extent to which they exhibit features of AI-generated language. Based on this, they identified a clear transparency gap: the use of AI appears to be significantly more common than what authors report.
The results suggest that current guidelines, which rely mainly on voluntary disclosure, are insufficient to guide the use of AI in academic writing. At the same time, they raise the question of how researchers' responsibility and the reliability of scientific publishing should be updated in an era when generative AI is rapidly becoming a commonplace tool.
Source: Academic journals' AI policies fail to curb the surge in AI-assisted academic writing, ArXiv (AI).
The study reviewed 5,114 scientific journals and over 5.2 million articles published in them. The majority of the journals – about 70 percent – had adopted some form of AI-related policy. Most often, the guidance was limited to asking authors to disclose if they have used generative AI, such as large language models, for producing or editing text.
However, the data shows that the use of AI is sharply increasing everywhere, and the difference between journals with AI policies and those without is not statistically significant. The growth is particularly rapid in countries where English is not the primary language, as well as in the natural sciences and journals where a large portion of the articles are openly accessible.
Researchers also conducted an in-depth text analysis of 164,000 scientific articles to assess the extent to which they exhibit features of AI-generated language. Based on this, they identified a clear transparency gap: the use of AI appears to be significantly more common than what authors report.
The results suggest that current guidelines, which rely mainly on voluntary disclosure, are insufficient to guide the use of AI in academic writing. At the same time, they raise the question of how researchers' responsibility and the reliability of scientific publishing should be updated in an era when generative AI is rapidly becoming a commonplace tool.
Source: Academic journals' AI policies fail to curb the surge in AI-assisted academic writing, ArXiv (AI).
This text was generated with AI assistance and may contain errors. Please verify details from the original source.
Original research: Academic journals' AI policies fail to curb the surge in AI-assisted academic writing
Publisher: ArXiv (AI)
Authors: Yongyuan He, Yi Bu
December 25, 2025
Read original →