From Community Values to AI Design Constraints: A Mixed-Methods Study of a Proposed AI-Driven Wildfire Risk Assessment Tool in Los Angeles County

Abstract

AI-driven hazard tools are increasingly being proposed for resident-facing risk communication. However, residents are often asked for feedback only after decisions about data use, privacy, and explanation have already been made. This study takes an earlier approach by examining residents’ values, concerns, and practical expectations related to a proposed AI-based risk assessment tool and its anticipated outputs before a predictive model or functional interface is developed. The proposed smartphone application would use exterior property photographs to estimate parcel-level wildfire risk and provide an interpretable risk score, uncertainty information, and mitigation recommendations. Using a convergent mixed-methods design, we surveyed 30 Los Angeles County residents, seven of whom also participated in a structured virtual town hall. The county was selected as the study setting because of its high wildfire exposure and recurring fire risk. We analyzed the data using descriptive statistics, exploratory FDR-adjusted Spearman correlations, and inductive thematic coding of open-ended responses and town hall discussion. Findings indicate that participants were cautiously receptive to the proposed tool: 66.7% said they would be very likely or likely to use the tool. They linked fair risk assessment to whether the tool considered relevant property and neighborhood conditions. Privacy and data security were the most common concerns (64.3%), while cost was the main barrier to acting on recommendations (65.5%). A strong association was found between participants’ likelihood of using the tool and their likelihood of saving or sharing their risk results (ρ = 0.752, pFDR < 0.001). We translated the combined findings into the AI Value Map, a community-grounded artifact organized around nine value dimensions. The Map connects participant findings to provisional design requirements and identifies possible consequences if those requirements are not addressed. This study contributes exploratory evidence on residents’ responses to a proposed AI wildfire tool and offers a pre-model procedure for translating early community input into traceable design constraints for resident-facing AI hazard tools.