With the increasing use of authentic materials in high school English reading instruction, adapting such texts to match students’ proficiency levels has become both essential and challenging. This study investigates the potential of ChatGPT in English reading text adaptation by comparing ChatGPT-adapted texts with expert-adapted texts based on ten sets of Gaokao reading materials (2022–2024). An analytical framework comprising four adaptation dimensions—lexical, syntactic, content, and cohesion—and 19 specific strategies was employed. The findings reveal that both experts and ChatGPT rely most heavily on content-level adaptation, while experts engage least in syntactic adaptation and ChatGPT in lexical adaptation. At the strategy level, experts tend to prioritize information omission and high- and low-frequency word adjustment, whereas ChatGPT shows a preference for explanation addition, paraphrasing, and reordering of information. ChatGPT demonstrates clear advantages in rapidly identifying grammatical errors and enhancing syntactic diversity; however, it also exhibits limitations, including insufficient awareness of test design conventions, over-adaptation of linguistic style, weak cohesion, and inadequate evaluation of informational relevance. These findings highlight the importance of teacher–AI collaboration in text adaptation. Teachers should assume the roles of leaders, evaluators, and optimizers to guide, assess, and refine AI-generated texts. This study contributes to understanding the application of generative AI in language education and provides practical implications for the development of reading materials.

Xinyu Hong. A Study on Generative AI-Assisted Adaptation Strategies for High School English Reading Texts. International Journal of New Developments in Education (2026), Vol. 8, Issue 7: 1-6. https://doi.org/10.25236/IJNDE.2026.080701.

[1] Zhang H Q, Gu X D. An analysis of methods for adapting English reading materials in the National College Entrance Examination: A case study of the 2018 National English Test[J]. China Examinations, 2019(11): 36–42.

[2] Young D J. Linguistic simplification of SL reading material: Effective instructional practice? [J]. The Modern Language Journal, 1999, 83 (3): 350-366.

[3] Çelik F, Yangın Ersanlı C, Arslanbay G. Does AI simplification of authentic blog texts improve reading comprehension, inferencing, and anxiety? A one-shot intervention in Turkish EFL context[J]. International Review of Research in Open and Distributed Learning, 2024, 25(3): 287-303.

[4] Crossley S A, Allen D, Mcnamara D S. Text simplification and comprehensible input: A case for an intuitive approach[J]. Language Teaching Research, 2012, 16 (1): 89-108.

[5] Rets I, Astruc L, Coughlan T & Stickler U. Approaches to simplifying academic texts in English: English teachers' views and practices [J]. English for Specific Purposes, 2022, 68: 31-46.

[6] Yuan W E, Li Z Y, Jin T. An evaluative study on machine-based text adaptation strategies for English reading materials: A case study of Gaokao reading corpus[J]. Foreign Language Testing and Teaching, 2023(04): 38–48.

[7] Bogaerds-Hazenberg S T, Evers-Vermeul J & van den Bergh H. A meta-analysis on the effects of text structure instruction on reading comprehension in the upper elementary grades [J]. Reading Research Quarterly, 2021, 56(3): 435-462.