ELT Go Getter Symposium 2025

Presenters: Kristin Terrill

Abstract: AI-proof homework assignments, AI detectors, and AI disclosure policies emerged almost immediately after OpenAI released ChatGPT, its web-based chatbot undergirded by the large language model (LLM) GPT-3, in November 2022. These counter-innovations reveal the threat perceived in academia by technology that can synthesize discourse. LLMs like ChatGPT and image and computer code generators have been collectively designated as generative AI (GenAI) and summarily inducted into the goon squad of cheating devices encompassing contract cheating, exam pre-knowledge, and plagiarism. Plagiarism is, of course, making a false claim of having written something original. In this talk I will show that although GenAI cheating is unethical, it is not plagiarism. Importantly, plagiarism necessitates appropriating a prior intellectual contribution by an author; because GenAI does not meet the definition of authorship, its misuse cannot be classified as plagiarism. A novel conceptual framework will be presented, connecting six elements of plagiarism to distinguish it from GenAI cheating in two target language use scenarios: homework and research article submission. The framework has utility as a teaching tool that reveals how plagiarism compromises the integrity of discourse and discourse communities. More importantly, it clarifies differences in values between contexts, highlighting the importance of assessment validity in the education context versus proprietarity of knowledge contributions in the scholarly publishing context.