AI-Generated Content Detectors: Boon or Bane for Scientific Writing

Abstract

Objectives: The purpose of the study is to help publishers identify AI-generated text in scientific research, academic works, and assignments as a critical step toward the regulation and promotion of the ethical usage of AI in academia. Method: Recently developed literature on Generative AI suggests that human reviewers may fail to distinguish between human and AI-generated articles. Therefore, the present study evaluates AI-powered software as a potential solution for AI-generated content detection. We performed an experiment to see whether AI detector tools are capable of identifying and distinguishing between human- and AI-generated texts. To determine the accuracy of the AI-detector, we created and submitted four research articles to AI-detector tools for a pre- and post- manipulation test. Findings: The study shows that it is quite impossible for any AI detector to identify all AI-generated content accurately, thus, human-AI collaboration strategies can be employed to achieve the maximum accuracy. This paper demonstrates in a novel manner how AI detector tools can be manipulated to provide false results. Novelty: To the best of the authors’ knowledge, this paper is the first to acknowledge growing AI-literacy among students and scholars. This study edifies academia by providing scientifically verified human-AI collaboration strategies to capitalize on these tools and thwart academic misconduct. Keywords: Artificial Intelligence, Chatbot, ChatGPT, Generative AI, Scientific Research

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 15 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.