This chapter explores GAN-based facial image editing, addressing tasks like expression, age, pose, style, makeup, and identity swapping. Expression editing uses keypoint-guided models (e.g., G2-GAN with cycle and identity loss). Age editing leverages latent space models (e.g., CAAE with adversarial autoencoders). Pose editing integrates 3DMM models (e.g., FFGAN and FaceID-GAN using face recognition features). Style transfer employs attention mechanisms (e.g., UGATIT with adaptive normalization). Makeup transfer (e.g., BeautyGAN) combines cycle consistency and histogram matching. Face swapping relies on geometric deformation or deep feature disentanglement. A unified framework (StyleGAN) enables attribute manipulation via latent vector editing. Finally, A very detailed practical project based on StyleGAN for face editing was demonstrated to help readers grasp the details.