---

The word “deepfake” was coined on a Reddit message board in late 2017. It originally referred to any video that had been altered using artificial intelligence (AI) to produce hyper-realistic depictions of people saying and doing things they never actually said or did. It has since become clear, however, that video is not the only medium vulnerable to AI-powered manipulation. In addition to deepfake videos, our world is now also awash in AI-generated text, audio, and images. Some observers prefer to draw distinctions among different types of synthetic media and different methods of production (i.e., generative AI, LLMs, GANs, diffusion models, cheapfakes, shallow fakes, etc.), but we have chosen a deliberately broad definition for deepfakes because we want a convenient label that encompasses any AI-generated content in any medium.

AI-generated synthetic media will not only outnumber traditional, human-generated content on the internet, but it will eventually outnumber all the media that we have ever created since the dawn of human culture thousands of years ago. Little wonder that people are freaking out. After all, who can you trust when you can’t even trust your own eyes? Many observers have expressed fear that deepfakes herald nothing less than the “information apocalypse,” the very “collapse of reality.” Rather than surrender to nihilism, we built this website to help educators keep it real in the deepfake era.

---