Have AI-Generated Texts from LLM Infiltrated the Realm of Scientific Writing? A Large-Scale Analysis of Preprint Platforms

Abstract

Since the release of ChatGPT in 2022, AI-generated texts have inevitably permeated various types of writing, sparking debates about the quality and quantity of content produced by such large language models (LLM). This study investigates a critical question: Have AI-generated texts from LLM infiltrated the realm of scientific writing, and if so, to what extent and in what setting? By analyzing a dataset comprised of preprint manuscripts uploaded to arXiv, bioRxiv, and medRxiv over the past two years, we confirmed and quantified the widespread influence of AI-generated texts in scientific publications using the latest LLM-text detection technique, the Binoculars LLM-detector. Further analyses with this tool reveal that: (1) the AI influence correlates with the trend of ChatGPT web searches; (2) it is widespread across many scientific domains but exhibits distinct impacts within them (highest: computer science, engineering sciences); (3) the influence varies with authors who had different language speaking backgrounds and geographic regions according to the location of their affiliations (>5%: Italy, China, average over countries), and (4) AI-generated texts are used in various content types in manuscripts (most significant: hypothesis formulation, conclusion summarization). Based on these findings, an AI-revision index is developed and calibrated, giving quantitative estimates about how AI is used in scientific writing. Suggestions about advantages and safe use of AI-augmented scientific writing are discussed based on our observations.