Understanding AI and Deepfakes In 1997, a Video Rewrite Program was published. It generated new videos of people mouthing words they didn't originally say by matching phonemes from audio tracks with mouth images from existing footage. [1] This marked the first recognized occurrence of a deepfake. Deepfakes, which are artificial audio, photo, or video manipulations, are a significant concern due to their heavy reliance on machine learning, thereby establishing a strong link between AI and deepfakes. These manipulations can convincingly impersonate individuals, making it difficult to discern the fake from the real. AI and Deepfakes Powered Cybercrimes The resulting cybercrimes are huge and disturbing because AI-created deepfakes involve manipulating cyberinfrastructure through cyber means in cyberspace. 1. Phishing Scams Phishing is the most common way of duping a person out of his money, and deepfakes have become increasingly unstoppable. Fraud in China used dee...