The advent of generative artificial intelligence (AI) technologies, such as large language models (LLMs) like GPT-4, has introduced a novel and escalating threat to academic publishing- the proliferation of AIgenerated fake papers. These fraudulent manuscripts, often produced by paper mills (Paper mills are a type of industrial fraud, which is prevalent in the publishing sector. Paper mills are profit-oriented, unofficial and potentially illegal organisations which produce and sell fabricated or manipulatedmanuscripts which resemble genuine legitmate research.) 1 or individual actors, undermine the integrity of scholarly communicaon by introducing fabricated data, plagiarized content, and hallucinated references into the literature. This paper explores the mechanisms behind this threat, its impacts on research credibility, and the limitations of current detection methods. Drawing on recent case studies and analyses, we highlight how AI exacerbates existing issues like publish-or-perish pressures (the academic and scientific pressure to continually publish research findings to advance one's career, secure funding, and achieve institutional recognition). Finally, we propose innovative approaches to mitigation, including AI-enhanced verificaon tools, blockchain-based authorship tracking, and collaborative industry-wide protocols. By addressing these challenges proactively, the academic community can safeguard the trustworthiness of scientific knowledge.
Chaudhuri et al. (Sun,) studied this question.