Measuring the Acceptance of AI-Assisted Design: Developing the AI-Assisted Design Acceptance Inventory through its Relationship with Technology Anxiety, Creative Self-Efficacy, and Aesthetic Sensitivity

Abstract

Abstract

Background and Research Gap: With the rapid proliferation of generative artificial intelligence (AI) in the field of design, designers are experiencing profound transformations in workflows and creative processes. From conceptual ideation to the generation of final prototypes, AI technologies are reshaping the entire design ecosystem. However, existing design research often overlooks the significant impact of individual differences among designers on the acceptance and effectiveness of AI tools. Although the technology itself continues to advance, its potential may remain underutilized without active user adoption. Both academia and industry currently lack standardized instruments to measure AI tool acceptance specifically in design contexts, limiting our understanding of the underlying mechanisms of human-AI co-creation.

Methods: This study systematically defines and operationalizes the core construct of AI-Assisted Design Acceptance (AI-ADA) and develops the AI-Assisted Design Acceptance Inventory (AI-ADAI). A rigorous survey was conducted with a large sample of design students and professionals (N = 494). Exploratory factor analysis (EFA) and comprehensive statistical tests were employed to evaluate the reliability and validity of the inventory.

Practical Approach: To examine the external validity and theoretical significance of the scale, multiple linear regression models were employed, incorporating three well-established psychological constructs: technology anxiety, creative self-efficacy, and aesthetic sensitivity. These analyses enabled in-depth cross-validation of the subdimensions of AI-ADAI and exploration of their underlying mechanisms.

Key Findings: Factor analysis revealed four latent dimensions of AI-ADAI: Human-AI Co-creation, Iterative Optimization, Ethical & Copyright Awareness, and Design Efficiency Enhancement. Regression results indicated that creative self-efficacy was positively associated with AI acceptance (β = 0.256, p < 0.001), whereas technology anxiety showed a significant negative relationship (β = -0.109, p < 0.001). Interestingly, aesthetic sensitivity did not exhibit a simple linear relationship with specific subdimensions (e.g., iterative optimization) but followed a complex nonlinear inverted-U pattern, providing new insights into experienced designers’ resistance to adopting AI technologies.

Significance: This study introduces the first rigorously validated standardized instrument for quantitatively assessing designers’ acceptance of AI tools. The findings enrich theoretical frameworks at the intersection of human-computer interaction and design, extend the applicability of traditional technology acceptance models, and offer targeted strategies for curriculum reform in design education and psychological interventions in organizational digital transformation.

Keywords: AI-assisted design; scale development; technology acceptance model; technology anxiety; creative self-efficacy; aesthetic sensitivity; human-AI co-creation