The thesis contributes to the study of Generative AI applications in digital marketing operations by proposing an AI-assisted validation framework designed for a real operational context, characterized by heterogeneous assets, legal and promotional constraints, brand consistency requirements, and the involvement of external suppliers. The main contribution lies in the definition of a controlled, traceable, and reusable pipeline able to integrate deterministic checks and generative model-based controls without removing the decision-making role of the human operator. From a methodological perspective, the thesis shows how a traditionally manual and fragmented process can be translated into a structured sequence of activities: file recognition, content extraction, retrieval of validated sources, application of validation rules, report generation, and human-in-the-loop review.