The work aims to study ways of using the artificial intelligence technology to assist in inspiring creativity from the perspective of design. To enhance the efficacy of creativity inspiration, 18 design students were recruited to use traditional search and Text to Image AI for user experience testing. Quantitative and qualitative data were collected through scales, interviews and non-participatory observation for pattern analysis to verify the effectiveness of artificial intelligence in assisting creativity inspiration. The results indicated that AI could effectively reduce cognitive load and support designers in their exploratory and expressive activities. However, despite reducing the workload, AI assistance did not significantly improve the novelty or quality of the design output, which was contrary to the subjective feelings of the participants. Furthermore, by analyzing the behavioral characteristics of designers under AI assistance, the designers were categorized into two types:Agentive and Collaborative. The Agentive users adopted the "Delegate-Adoption" Pattern, which was manifested as "Commissioning - Accepting - Adopting"; and the Collaborative users adopted the "Assistive-Iterative" Pattern, which was manifested as "Dismantling - Rejecting - Iterating". Through the user experiment test, this research analyzes the capability of artificial intelligence in assisting creativity inspiration, and summarizes the behavioral logic and reasons for these classifications, providing insights for subsequent studies on integrating AI into traditional design workflows.