Deepfake Professionals

In 2014, a doctoral student named Ian Goodfellow (who is now a research scientist at Google DeepMind) invented generative adversarial networks (GANs), a machine learning technology to expand the potential uses of artificial intelligence (AI). Although GANs could be used to create positive content, “bad actors” began to use them to superimpose the faces of famous actresses and other public figures into pornographic videos. In 2017, a Reddit user named “Deepfakes” uploaded well-designed pornographic videos that featured these types of fakes, and the deepfake (as it was named) phenomena took off. It’s estimated that 98 percent of deep fakes are pornographic. The use of deepfakes to discredit public figures, manipulate public opinion, and commit crimes has become common. In December 2018, there were an estimated 7,964 non-consensual and harmful deepfakes, according to Sensity, which bills itself as the world’s first visual threat intelligence company. The total number of deepfakes increased to 95,820 in 2023. Although there are many positive uses for deepfakes, government officials, computer security specialists, computer ethicists, and the public have serious concerns about their negative and even criminal usage. In 2019, the DeepTrust Alliance was founded to “advance the fight against digital disinformation and deepfakes.”