This paper explores the role of prompt-based generative AI in artists' creative practice through a diffractive analysis. The study involves seven visual artists who were provided with a personalized instance of Stable Diffusion, fine-tuned on their artwork. The analysis identifies two dominant modes of use: AI for ideation and AI for production. Participants found the personalized model more engaging, particularly for generating novel ideas and exploring their artistic style. However, they also expressed concerns about plagiarism, efficiency, and the impact on their creative identity. The paper concludes with design considerations for future generative AI interfaces, emphasizing the need for a balanced approach that respects human creativity and addresses ethical implications.This paper explores the role of prompt-based generative AI in artists' creative practice through a diffractive analysis. The study involves seven visual artists who were provided with a personalized instance of Stable Diffusion, fine-tuned on their artwork. The analysis identifies two dominant modes of use: AI for ideation and AI for production. Participants found the personalized model more engaging, particularly for generating novel ideas and exploring their artistic style. However, they also expressed concerns about plagiarism, efficiency, and the impact on their creative identity. The paper concludes with design considerations for future generative AI interfaces, emphasizing the need for a balanced approach that respects human creativity and addresses ethical implications.