ecoAIware: Exploring Public Awareness of Generative AI’s Environmental Footprint Through Participatory Data Physicalizations
Anniek Jansen, Supraja Sankaran · 2025
Generative AI is becoming increasingly accessible to non-expert users. However, these technologies consume substantial amounts of environmental resources, often sourced non-sustainably. This environmental impact remains largely invisible to users, and it is unclear how aware the public is of these environmental costs. The intangible nature of generative AI, like other cloud-based technologies, suggests that awareness of these hidden environmental impacts may be limited. To explore and increase public understanding, we propose ecoAIware, an interactive demonstration using participatory data physicalizations (PDPs). These PDPs represent electricity and water consumption, CO2 emissions, and individuals’ perceptions of the importance of the societal and environmental impacts of generative AI. By embodying these metrics, we aim to engage and inform users in a tangible, interactive way. ecoAIware can spark critical reflection and discussion while also serving as a tool for gathering data on public awareness, thus contributing to the growing discourse on generative AI’s environmental footprint.