Generative Adversarial Network or GAN is an unsupervised Deep Learning Neural Network model introduced by Ian Goodfellow in 2014. A GAN consists of a Generator and a Discriminator, The Generator learns to generate new data with the same statistics as the training set, while the Discriminator tries to identify whether the data is real or fake. They contest with each other as a zero-sum game, where one agent's gain is another agent's loss.
As the generator and discriminator train against each other, they get better and better at their respective tasks. The generator learns to generate more realistic data, and the discriminator learns to better distinguish between real and generated data. Eventually, the generator can produce data that is indistinguishable from real data to the discriminator.
Research paper for more reference.