As the frequency and severity of climate-related disasters continue to rise, large language models (LLMs) have the potential to address inequalities in disaster response by enabling real-time data processing, multilingual communication, and improved accessibility for vulnerable populations. These models can facilitate the more equitable distribution of disaster information across diverse communities and provide tailored information to marginalized groups, such as people with disabilities and migrant workers. However, critical challenges remain, including algorithmic bias, the digital divide, and data privacy concerns. This raises important questions about whether LLMs can contribute to building a more inclusive disaster response framework or if they risk exacerbating existing vulnerabilities. Dr. Park Eun-seon, director of Listen to the City and an expert in disaster studies, will present recent research on the use of LLMs in disaster response, while team member Cho Ye-jin will lead a discussion on AI and ethics. In particular, Dr. Park will share insights from her hands-on experience in collecting data and developing an LLM-powered fake news detection model for the recent floods in Spain. The session will explore the background of the model’s development and the challenges encountered in the process. Participants are encouraged to critically reflect on the role of AI in disaster response and engage in a collaborative discussion to share ideas and insights.

Listen to the City is an art, urban, architecture, design, and research collective that began in 2009, dedicated to making the unseen presence within the city visible. For us, aesthetics means listening to the voices of the invisible and the marginalized and marking their place on the map.

Park Eun-sun studied fine arts and urban engineering and earned a Ph.D. in disaster management within the field of urban engineering. Team member Cho Ye-jin has been conducting research in media, philosophy, and anthropology.