SpheriCity: Designing Trustworthy Conversational AI for Sustainability Decision Support
Researchers have developed SpheriCity, a conversational AI prototype designed to enhance trust and transparency when analyzing sustainability reports. The system addresses challenges posed by large language models, such as opaque reasoning and hallucinations, by prioritizing evidence traceability and structured synthesis. An expert review indicated that transparent sourcing, contextual explanations, and integration into existing workflows are crucial for building user trust in AI for high-stakes decision-making in sustainability. AI
IMPACT This research offers a framework for building more trustworthy AI tools in critical domains like sustainability, potentially improving decision-making processes.