STEAM art course design combining generative AI and prompt engineering

This study proposes a dual-driven STEAM art course framework integrating generative AI and prompt engineering to address key challenges in creativity support, personalisation, and AI tool integration.A four-layer topology aligns course goals, content, interaction, and evaluation, combining generative AI's creative abilities with prompt engineering's precision.Hierarchical prompt strategies enable stepwise creative guidance, while a course-AI feedback loop adapts to learner needs.A multidimensional evaluation system assesses creative expression, skill development, and thinking growth.Results show a 42.3% increase in creative work scores, 91.7% skill proficiency, 4.8 satisfaction (out of 5), and 35.6% higher teaching efficiency.Personalised teaching coverage rose from 38% to 89%.The framework performs effectively across diverse age groups and skill levels, offering a scalable path for intelligent art education in K-12 and training contexts.