LLMs and Generative AI for the Detection and Generation of Scientific Content
El Hari Karima, Soussi Ilham · 2025
Generative Artificial Intelligence (GenAI) in the form of Large Language Models (LLMs) has revolutionized the way that scientists interact with scholarly writing. They enable the creation of quality texts, making plagiarism detection a crucial field for human beings. This progress involves the creation of a powered Artificial Intelligence (AI) detection methods. Additionally, the use of LLMs to scientific content creation by helping in manuscript writing, research summaries and even generating hypotheses. AI application in both stages of generation and verification of scholarly material requires a methodical examination of available methods and limitations in this field. This research carries out a Systematic Literature Review of 58 studies to examine the methods, datasets and evaluation techniques applied in their integration within scientific writing. The results list ways of using LLMs in academic research like framework development, as well as the need of overcoming challenges such as hallucinations. The findings highlight the demand for stronger systems of evaluation, better AI detection software and more detailed ethical guidelines for responsible use of LLMs within the academic field. This review is contributing to the discussions of the future use of AI in academic communication, with descriptions of how LLMs are capable be effectively integrated while minimizing potential risks.