시민과학 기반 AI 거버넌스에 대하여

초록

This paper proposes governance that applies the methodology of citizen science to practically construct and enforce AI ethics. The development of AI governance requires new methods that reflect the characteristics of AI technology. Since AI technology evolves continuously through learning, the stages of development and use are not distinctly separated. While this presents a technological risk on the one hand, it also allows users to engage with the technology on the other. Recently, issues surrounding AI technology have shifted into matters concerning public ways of life. Therefore, we seek to explore a model of “AI governance for learning and adaptation” that considers both the technical characteristics of AI and user participation. First, the paper presents the basic principles for designing AI governance through a theoretical review. Then, it analyzes three AI cases. Synthesizing these preliminary analyses, it constructs a “citizen science-based AI governance”. Data collectively generated by citizens can provide the foundation for practical AI governance, where AI and citizens can learn and adapt together.

키워드