Documentation and Transparency in AI-Supported Research

Abstract

This study offers a comprehensive examination of the ethical dimensions related to the use of artificial intelligence (AI) in scientific research, especially as reliance grows on generative AI tools like ChatGPT and Claude. The study begins with the Philosophical question: “Who wrote the research – the human or the AI?” and explores four main dimensions: ethical boundaries, intellectual property, institutional policies, and transparency in documentation.

Findings revealed a growing consensus that using AI without disclosure, or for generating core research elements, constitutes an ethical violation. The study also confirmed that AI cannot be credited as an author, though its role must be acknowledged when it contributes significantly. The results highlighted discrepancies in institutional policies and practices across disciplines, underscoring the need for unified standards. The study proposed a three-part framework to assess the ethical use of AI: task appropriateness, level of human oversight, and degree of transparency.

It recommended flexible, principle-based policies, ethical training for researchers, and precise documentation practices. Overall, the study advocates for a responsible partnership between humans and AI that safeguards scientific integrity while embracing technological advancement.