This document discusses deepfakes, including what they are, their history, present uses, future challenges, and consequences. Deepfakes use deep learning techniques like GANs to manipulate images and audio to deceive viewers into thinking something is real when it is actually fake. While initially developed by researchers, open-source tools now allow anyone to generate deepfakes. The future poses challenges around reducing training data needs, improving temporal coherence in videos, and preventing identity leakage, among other issues. Deepfakes could potentially target politicians, actors and public figures to manipulate perceptions. Prevention strategies include developing counter-AI techniques, using blockchain, and raising awareness.