Elouise Botes is a postdoctoral researcher in Psychology at the University of Luxembourg, where she is affiliated with the Institute of Cognitive Science and Assessment and the Luxembourg Centre for Educational Testing. She is an inter-disciplinary researcher examining the role of individual differences in language learning, using concepts from both psychology and applied linguistics in her work. She was recently awarded with the Early Career Researcher Award from the International Association for the Psychology of Language Learning and is currently on the editorial board for the Journal of Multicultural and Multilingual Development.
Responsible Linguistics in the Age of Generative AI: Implications for Researchers, Reviewers, and Editors
Generative AI and large language models (LLMs), such as ChatGPT, are reshaping academic writing, with growing evidence of their integration into linguistics and applied linguistics research. Drawing on recent analyses of over 26,000 abstracts from the top 100 linguistics journals (2020–2024), I demonstrate a significant rise in the frequency of stylistic markers consistently linked to LLM-generated text, emerging sharply in 2024. These patterns suggest that LLM use is no longer peripheral, researchers are using LLMs to write blocks of texts in research publications. This keynote will examine the ethical, practical, and epistemological implications of these shifts for the linguistics community. For researchers, LLMs raise questions of authorship, originality, and accountability. For reviewers, risks include bias, hallucination, and manipulation of AI-assisted peer review. For editors, the need for clear, enforceable policies is urgent. I will outline current ethical frameworks for LLM use by researchers, discuss detection approaches and their limitations, and the coming impact of EU AI regulations. Ultimately, as a field, we need to find a path forward for responsible engagement with LLMs that preserves scholarly integrity while acknowledging the realities of a rapidly evolving technological landscape.