<p>Artificial intelligence is increasingly present in everyday UX design work, particularly during early stages such as ideation and interface structuring. This study explores how collaboration between human designers and AI systems shapes the UX design process, with attention to interaction quality and sustainability at the level of design practice. A comparative experiment was conducted with two groups of designers: one working without AI support and another incorporating AI-assisted tools during early design activities. Rather than treating AI as a single or uniform technology, this research focuses on a class of generative, layout-oriented systems commonly used in UX prototyping. Across several controlled design scenarios, differences emerged not only in the amount of design effort required, but also in how interaction structures evolved. AI-supported workflows tended to converge on stable layouts earlier and required fewer structural revisions, which in turn affected how users navigated multi-step tasks during usability testing. From a sustainability perspective, this study adopts a process-oriented view of sustainable UX, defining it in terms of reduced redundant effort, lower cognitive load, and the ability to maintain and adapt interaction designs over time. The findings suggest that sustainability in UX may arise less from the tools themselves than from how AI support reshapes the distribution of design effort within human-centered workflows. By examining these dynamics, this research offers empirical insight into human–AI synergy as it appears in practical UX design, contributing to ongoing discussions about how AI can support more sustainable interaction design practices without replacing human judgment.</p>