This survey consolidates and analyzes recent developments in using Electroencephalography (EEG) signals to generate images, text, and audio with generative AI. The study finds that EEG-to-image models use encoder-decoder architectures, EEG-to-text approaches use transformer-based language models, and EEG-to-audio methods map EEG signals to mel-spectrograms. The survey highlights the challenges in the field, including small and heterogeneous datasets, limited cross-subject generalization, and the absence of standardized benchmarks. It provides a foundational reference for advancing EEG-based generative AI and offers open-source datasets and baseline implementations for systematic benchmarking.