A Feature Mapping on GenAI Tools from the Perspective of HCI
摘要
This paper presents a feature mapping-based evaluation of generative artificial intelligence (GenAI) tools from a Human-Computer Interaction (HCI) perspective. This mapping aims to explore and compare how GenAI tools support interaction, usability, and user-centered features. This is relevant because GenAI systems become increasingly integrated into creative, educational, and productivity contexts. For this mapping, we chose the tools developed and provided by the company responsible for the base model, as cited in the AI Index Report 2025 and with high popularity in 2024, according to the AI Tools portal. We analyzed 15 publicly available GenAI platforms, using a multidimensional framework that considers interaction styles, interface types, and functional aspects such as transparency, collaboration, user support, adaptability, and technical accessibility. The results reveal strong convergence in interface design, with 100% of tools adopting graphical interfaces and 93% supporting touch. However, advanced features such as multimodal integration, response customization, real-time feedback, and collaborative editing are unevenly distributed across platforms. Only 67% of tools support actual multimodal interaction, and just one tool offers real-time collaborative editing. In addition, transparency about data usage and the presence of explanations or citations in outputs remains limited. This mapping details the maturity of GenAI tools regarding interaction design and highlights areas where HCI principles remain underutilized. The findings support future research and development focused on enhancing user experience, inclusiveness, and contextual adaptability in emerging GenAI systems.