Abstract

Generative AI acquires knowledge and interacts with people in a human-like manner through multimodal interfaces and generates probable answers by understanding relationships among vast amounts of data. As a result, generative AI, once confined to laboratories, is now being recognized as a commercialized service, transitioning from a technology accessible only to a few experts to one that is easily accessible to users at various levels. The human-like ability of generative AI technology to acquire knowledge and interact, along with its autonomous learning capability, is expected to have significant ramifications across our socio-economic spectrum. This study discusses the potential of generative AI and the government's role based on expert surveys. It is composed of four parts. Firstly, it presents the key features demonstrating the potential of generative AI, followed by an analysis of the evolutionary direction, ramifications, and future policy challenges. Lastly, it suggests the AI policy direction in terms of ① securing AI technological competitiveness and regulatory framework, ② activating the AI ecosystem, ③ establishing AI governance, and ④ accepting AI technology and adapting to changes.