Designing Discursive Identity: Integrating Indigenous Relational Values with Informational Baselines in Public-Facing Environmental Chatbots

This study examines how assigning a specific cultural identity to an AI changes how it communicates complex environmental topics. We compared two public-facing water conservation chatbots: Waterbot, a standard baseline providing objective scientific facts, and Riverbot, which integrates Indigenous relational perspectives by answering from the perspective of a river. By analyzing user interactions with the bots, we found systematic differences between the two models. Riverbot generated responses at a more accessible reading level than the Waterbot baseline. Additionally, while Waterbot remained emotionally neutral, Riverbot expressed a broader range of nuanced emotions, functioning as a more relatable conversational agent. These findings indicate that public-facing AI design should go beyond simple information delivery. Integrating relational and cultural values into a chatbot’s identity can improve its emotional resonance and accessibility, offering a practical approach for designing future environmental communication tools.