This chapter explores the application of generative adversarial networks (GANs) for synthesizing brain magnetic resonance imaging sequences in the context of multiple sclerosis (MS). It presents advanced MRI synthesis methods, including lesion-focused loss functions for improved lesion appearance and uncertainty quantification in synthetic images. It details the technical aspects of GANs, including their architecture, training, and optimization, and discusses their clinical applications from diagnostic enhancements to their integration into multi-center studies. Furthermore, the chapter assesses the validation of these models in clinical settings, showcasing their ability to enhance diagnostic accuracy, detecting and monitoring MS. Through extensive experiments and reader studies with experienced radiologists, it was demonstrated that synthetic images achieve high-quality clinical utility. Finally, the chapter discusses the limitations and future directions of generative MRI synthesis in MS, highlighting its potential to impact clinical practice and patient care.