“I didn’t say that!” …The world will be forever changed by Deepfakes. A portmanteau of “deep learning” and “fake”, this trend refers to a new AI-assisted human image synthesis technique that generates realistic video face-swaps which can even be done in real-time. With current video technology, a collection of still selfies can be used to create realistic videos. Fake is the new reality: A voice can be faked. A face can be faked. What happens when we cannot trust what we hear and what we see? Is it possible to detect and defeat this global threat? The challenges from faked content emphasize the need for society remaining critical to protect ourselves as this technology gets better and better, but like any technology, the AI which creates deepfakes is neither good nor evil. This talk explores the possible ramifications and ethics of deepfakes, from privacy and consent, to First Amendment and parody rights, to global thermo-nuclear war, and considers both positive and negative possible scenarios.