Abstract

The rapid diffusion of generative AI requires science museums and other knowledge institutions to design AI services that preserve the accuracy and context of their collections.
Focusing on the heritage AI service HAI, this presentation examines domain-specific data, RAG-based reliability, generative image services, data control, and operational trade-offs.
Drawing on HAI’s selection and introduction as an international case at the AI for Good Global Summit 2026, the presentation offers practical implications for the design and governance of AI services in science museums.

Biography

· 2006–Present, Head of the AI Data Team, Korea Heritage Agency
· 2023–2024, Completed AeG Program (Served as Secretary General), KAIST
· 2008, Completed ITCEO Program, Soongsil University
· 2003–2004, CEO, WAAYO Corporation