ISACA AI Fundamentals Practice Test 2026 - Free AI Fundamentals Practice Questions and Study Guide

In image generation, the model trained through adversarial competition between two networks is Generative Adversarial Networks. It consists of a generator that creates images from random inputs and a discriminator that tries to tell real images from those generated. They train together in a minimax game: the generator aims to fool the discriminator by making increasingly realistic images, while the discriminator gets better at detecting fakes. This back-and-forth pushes the generator to learn the true distribution of real images, producing high-quality samples. Other options don’t fit because a transformer or LSTM are designed for sequence data and attention or temporal patterns, not this two-network adversarial setup, and data collection with preprocessing is just data preparation, not a generative model.