Driving User Acceptance of AI-Enhanced Weather Applications: A Behavioral Study Using UTAUT2 and Trust

Weather applications are widely used to support daily activities, yet users continue to face challenges related to trust, personalization, and overall engagement. Artificial intelligence (AI) o ff ers opportunities to improve both usability and forecast accuracy in weather apps, particularly through features such as chatbots, contextual recommendations, and personalized weather information. Despite these advancements, there is limited empirical understanding of how AI-based features influence user acceptance compared to traditional weather applications. This study evaluates user acceptance of a weather app prototype enhanced with AI capabilities, using the Unified Theory of Acceptance and Use of Technology 2 (UTAUT2) as the primary framework, extended by the Trust construct. A quantitative and qualitative approach was employed through a questionnaire that aims to identify whether AI features can enhance user satisfaction, and which features are most valuable for future development of weather applications. A mixed-method approach was employed involving a comparative questionnaire completed by 101 participants, assessing both a generic weather app and the AI-enhanced prototype. The results reveal that the AI-based prototype significantly outperforms generic apps in several UTAUT2 constructs, including Performance Expectancy, Trust, Hedonic Motivation, Habit, and Social Influence. Qualitative feedback emphasizes the value of AI-driven personalization (e.g., clothing suggestions, activity planning), real-time accuracy, and calendar integration. These findings suggest that AI capabilities can meaningfully improve user satisfaction, trust, and behavioral intention to use weather apps. The study provides key insights for future design and development of intelligent, user-centered weather applications.