Generative AI has become a transformative tool in advertising and marketing, allowing for the automated creation of targeted and engaging marketing content. However, ensuring that AI-generated descriptions are coherent, meaningful, and resonate with target audiences remains a key challenge for marketers. This study addresses this challenge by using BERT embeddings to analyze and cluster a dataset of product descriptions intended for advertising purposes. Our methodology involves leveraging BERT to generate embeddings that capture the semantic meaning of each description, followed by K-Means clustering to identify common themes in advertising language. The clustering results reveal distinct marketing themes, such as luxury, innovation, and value, which align closely with traditional advertising strategies and brand messaging. This approach provides a systematic way for marketers to analyze and understand AI-generated content, making it easier to assess whether it aligns with brand objectives and consumer expectations. The findings suggest that generative AI, when combined with sophisticated analysis tools, can be a valuable asset in automating and enhancing marketing content, enabling scalable and effective advertising strategies.