The rapid diffusion of Generative Artificial Intelligence (GenAI) is deeply transforming knowledge‑intensive work, thanks to its ability not only to automate repetitive tasks but also, and more importantly, to enhance workers’ capabilities for creating new knowledge. This is particularly evident in creative sectors where novelty is critical and new knowledge is continuously generated to provide fresh stimuli to the market daily, as in the fashion industry. Despite this, limited empirical research has examined how marketing professionals integrate GenAI into their work. This study addresses this gap by investigating the factors that affect the adoption and use of GenAI among fashion marketing professionals, who are tasked daily with generating innovative ideas to engage potential customers and guide them towards purchasing their company’s products. To this aim, the study applies the Diffusion of Innovation Theory as a conceptual interpretative lens to examine the integration of GenAI among 30 purposely selected fashion marketing professionals. This theory, one of the frameworks that current research is using to investigate GenAI adoption processes, assumes that the diffusion of a new technology is influenced by five factors: relative advantage, compatibility, complexity, trialability, and observability. Data were gathered through semi-structured interviews and thematically analysed using MAXQDA software. The analytical process used a hybrid coding strategy, in which the five theory factors provided the deductive structure, while inductive themes were developed to capture the specific knowledge implications of human-AI collaboration. The findings of the investigation shed light on the current usage, benefits, and challenges of employing GenAI to support the knowledge-intensive work of fashion marketing experts. From an academic perspective, it confirms the usefulness of the Diffusion of Innovation framework for investigating the adoption of this new technology. At the same time, it offers valuable insights for marketing experts on the use of GenAI to support their professional activities. The usual limitations of interview-based studies regarding the generalizability of results are acknowledged.