In this paper, we investigate the possibility of leveraging the predictive power of machine learning to generate animated lightning bolts in the image space efficiently. To this end, we selected state-of-the-art machine learning architectures based on Generative Adversarial Network (GAN) and trained them on the commonly available videos. We demonstrate that visually convincing animations are achievable even when employing a limited dataset. The visual realism of the generated sequences of lightning bolts is assessed by conducting a user study on the participants.