Year: 2026
Type: conference-paper
Abstract: Text generated by large language models (LLMs) has been shown to be stylometrically distinct from human-written text (André et al., 2023;Shah et al., 2023;Opara, 2024;Soto et al., 2024;Li and Zhang, 2025;Selvioglu et al., 2025).But LLMs are increasingly used not only to generate text but also to edit human writing, and it is unclear whether the two leave the same trace.We show that AI generation leaves a consistent "stylometric footprint": a small subset of features, primarily entropy and lexical diversity, consistently separates AIgenerated text from human writing across 8 LLMs and 5 domains, while the remaining features depend heavily on the domain and generator.AI editing, however, does not reproduce the same footprint.Relative to their humanwritten sources, AI-edited texts show only a small increase in lexical diversity and a decrease in entropy, rather than the joint increase that characterizes AI generation.Lexical density, which contributes little to generation, instead becomes the dominant editing-associated signal.Stylometric features therefore separate AI-edited text from AI-generated text but are substantially less effective at separating it from human-written text.Our results suggest that "AI text" is not a single phenomenon: generation and editing leave qualitatively different stylometric traces and should be studied separately.
Authors: Zhengyang Shan, Yukyung Lee, Sophie Hao
Institution: Boston University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W7204811216