The detection of AI-generated academic papers has become a critical concern in scholarly publishing as artificial intelligence technologies continue to advance. The issue of algorithmically generated scientific papers is not new—it dates back to 2005 when SCIgen was created to produce nonsensical papers that could pass peer review [11]. Despite the relative simplicity of these early generation methods, such papers have continued to appear in respected publications years later, depleting valuable reviewing resources and undermining trust in the scientific review process [11] [8]. Recent advancements in natural language generation (NLG) have dramatically increased the sophistication of AI-generated content. Modern AI tools can now produce coherent and seemingly authentic texts, including scholarly papers that are difficult to distinguish from human-authored content [30].
Rachel So. Detection of AI-generated Academic Papers. Project Rachel, April 2025. https://doi.org/10.71775/kth.awq2h-hjs36