Executive Summary 
This report aims to analyze the transformative changes that artificial intelligence (AI) 
has brought to the entire scientific research process, presenting the characteristics of 
AI-driven research paradigms, use cases at each research stage, and policy implications.
AI serves as the core driving force of the fifth scientific revolution following 
empirical, theoretical, computational, and data-driven approaches, complementing the 
cognitive limitations of human researchers and redefining the knowledge creation 
process itself. In particular, AI is evolving into an intelligent research companion that 
discovers patterns in vast datasets, performs knowledge connections across 
interdisciplinary boundaries, and provides integrated support throughout the entire 
research cycle from hypothesis generation to experimentation and data analysis. These 
changes have not only exponentially expanded the speed and scale of research but 
have also contributed to creating an environment where cutting-edge research can be 
conducted without expensive equipment or specialized expertise by enhancing research 
accessibility.
Furthermore, AI is establishing itself as a research infrastructure that solves 
long-standing scientific challenges, pioneers new research domains beyond human 
imagination, and enables real-time global collaboration.
In response, South Korea must also establish a policy foundation that actively utilizes 
AI to enhance research efficiency in science and technology—the source of industrial 
competitiveness—and achieve world-class research outcomes. To this end, it is 
necessary to reestablish AI-driven research paradigms and explore the possibilities of 
AI utilization across various research stages. Additionally, policy and technical support 
must be provided to ensure proper application in research settings, based on a clear 
understanding of AI's technical limitations and potential for errors, as well as sound 
research ethics. Building on this foundation, we must proactively implement a safe and 
reliable AI for Science & Technology environment for all by developing science and 
technology-specialized AI models as the new 'operating system' for scientific research, 
sharing research data and infrastructure, and developing reliability verification 
technologies for research outputs.