Defining DeepFakes
In its most basic terms, a DeepFake is simply the combination of deep learning and a fake video. At the basic level, these "videos" are produced by taking in enough imagery, be it other footage or simply pictures of a target, and then using applied deep learning algorithms to stich the images together into a believable piece of content. With enough available media, it is very possible to create a near-Hollywood quality video of any person saying or doing anything. This ability tailors itself well to disinformation and fake news campaigns and, if paired with social media attack vectors, can exponentially increase the threat that is posed by this type of attack.
In yet another twist on the old paradigm of "no good deed goes unpunished," the technologies that have been the precursor for DeepFakes came from the very tools that have been used to create the content that we all enjoy today. Tools like Adobe Photoshop and other video editing software...
