AI in the Hands of UX Designers : Where and How UX Designers Best Can Integrate AI in the Design Process

Sammanfattning

Artificial intelligence (AI) offers new opportunities to enhance user experience (UX) design workflows, but how to effectively integrate AI across different design stages remains much to be explored. This master’s thesis examines the integration of AI into the UX design process, applying the Double Diamond framework as a guiding framework. A case study approach was applied using the Design Thinking methodology, involving semi-structured stakeholder interviews with UX designers and a development manager. The interview data from the stakeholder interviews were analyzed through thematic analysis to create themes and identify the core issues. Based on the identified needs, a brainstorming session was conducted with one of the UX designers from the case study to generate potential solutions. This led to the development of a prototype tool, the “Lo-Fi Generator,” designed to automate low-fidelity interface prototyping using components from an existing Geographic Information System (GIS) project, a platform used for mapping and analyzing spatial data for infrastructure planning, chosen for its relevance to the UX designers’ workflow. The prototype was evaluated through task-based usability testing, the System Usability Scale (SUS), and qualitative follow-up questions. The findings indicate that the UX designers perceive AI as most beneficial in the Develop phase of the Double Diamond model, where AI can assist ideation and prototyping, especially under time constraints. Participants viewed AI as a supportive tool to boost designer creativity and efficiency rather than a replacement for the designer, though they acknowledged challenges such as maintaining human oversight and ensuring data quality. The Lo-Fi Generator prototype received a high SUS score (83), reflecting good usability, and the UX designers expressed interest in using such a tool to generate initial design concepts. Its usefulness was also shown to depend on the project, as it was seen as more useful in new projects, where the Lo-Fi Generator can help in exploratory work, such as brainstorming and sketching, whereas in ongoing projects, it is more often task-oriented, with adjustments and refinement work, where the Lo-Fi Generator was seen as not as helpful. By automating repetitive tasks and expanding the range of design alternatives in these contexts, AI has the potential to enhance both efficiency and creativity. However, successful integration requires addressing UX designers’ concerns and positioning AI as a collaborative partner rather than a replacement for human creativity.