The term "deepfake'' is a combination of the terms "deep learning" (an AI-based learning system) and "fake." Deepfake refers to the creation of falsified content that appears credible and authentic due to artificial intelligence. The 2019 report from DeepTraceLab on deepfakes reveals that there are 14,678 of them, 96% of which are sexually explicit. These pornographic deepfakes often involve superimposing one person's face onto another's body and have harmful repercussions, predominantly affecting women and girls (Maddock, 2020). While the creation of fake pornographic videos may be seen as a form of creativity and entertainment in the pornography industry, this review focuses on the non-consensual creation and dissemination of synthetic pornographic images where consent from  involved parties is violated (Alilunas, 2024). These videos, though artificial, have very real emotional and social impacts (e.g., dignity infringement, false sexual identity, privacy violation), thus constituting a new form of technology-facilitated sexual abuse (Burkell et al., 2019). This practice is now recognized as a form of cyber violence against women and girls, sometimes used for political purposes to silence those who publicly express critical views, such as Rana Ayyub¹ et Bella Thorne² (Maddock, 2020). This review thus presents the ethical, legal, and social implications surrounding the non-consensual creation and dissemination of deepfakes on the internet.

¹https://www.nytimes.com/2018/05/22/opinion/india-journalists-slut-shaming-rape.html

²https://www.bbc.com/reel/video/p07r9vbk/the-real-and-fake-sex-lives-of-bella-thorne