Predicting intent in textual data has become increasingly important with the proliferation of conversational environments and advanced dialogue systems. Accurately determining the intent behind texts is crucial in these contexts. Recent advancements leverage deep-learning techniques, including BERT, GAN-BERT, and the RASA AI framework, which have shown significant improvements, especially in overcoming the challenges posed by limited, diverse datasets. Here, a novel deep-learning-based approach is proposed that enhances intent prediction in textual data by incorporating the benefits of Generative Adversarial Networks (GANs). Specifically, we integrate the RASA AI framework with the GAN-BERT model and evaluate its performance on two well-known intent classification datasets. Our results demonstrate the effectiveness of this combined approach in improving intent prediction accuracy.