By Hunfa Khan
@redlineproject
AI disclosure: ChatGPT was used to generate ideas for the headings and a basic outline for this post. It was also used to create the summary and FAQ sections. Hailuo.ai was used to create the image, and Google NotebookLM was used to create the podcast and video summaries. Read more about our AI use on our principles page.
Summary: This essay examines how AI technologies create deepfake images and how they are misused for exploitation. It highlights the ethical, legal and social consequences, particularly for women. The consequences of non-consensual deepfakes advocate for stronger laws, accountability, and ethical failures to protect victims and prevent further digital abuse.
How AI has Changed the Trajectory of Our Lives
AI has been around for many years without us realizing. From web searches to targeted ads, you name it. AI has held a big role in shaping today’s society. Although AI has existed for many years, ordinary people were not able to use it back then, as they can now. It has become accessible to students, faculty, elders, non-educators and anyone with internet access.
This is scary yet revolutionary. Over the past few years, a lot of people have been making use of this tool in their daily lives. Some are talking to it as a friend and some as a therapist. Besides AI being a powerful tool and helping people in their daily lives, there are many drawbacks to it. One of the most rapidly growing drawbacks is people altering existing images and then using them to blackmail victims, especially women.
At first, it was easy to distinguish an AI image from a real image, but as these AI models are training every day, 24/7, to become better at their tasks, it has become harder to note the difference sometimes. Many people have been fooled by the rapid use of deepfakes and images/videos curated with AI.
I used to find it humorous how the elderly family member would believe these pictures, as it was a new concept to their generation. However, the realistic nature of these images is becoming the new norm for our generation. Some adults are using this tool as a means to extort from youngsters, women and sometimes men, too. The rapid misuse of this tool has led many individuals to be distressed.
According to a 2025 article by “The Conversation,” individuals have also taken AI images/videos of children being sexually abused and altered them to put on the dark web. It was appalling to be informed about women going through this, but more horrific to hear about children being victims of these gruesome acts. In this essay, we will explore how AI deepfakes are affecting today’s society and are being used in an unethical manner. While AI-generated images can be used for ethical means, they are also being used in a non-consensual manner to extort and defame women.
The Process Behind the Creation of Deep-Fake Images
Artificial intelligence tools are actively training on and have been trained using real human faces and interactions. They have been caught various times for stealing images of real humans and creating their deepfakes. New versions of AI, such as nano-banana, DALLE, Google Gemini and Midjourney, have improved their models to a greater extent over the last few years by training on real images stolen by their respective owners.
According to a 2022 study by Hany Farid, a generative adversarial network (GAN) creates new images but is adversarial because it has two parts to it: A generator that makes the fake images and a discriminator that determines if the image produced is real or fake. For instance, when the generator is asked to create an image, it generates a random image, and the discriminator attempts to check its database to see if the presented image matches a real human. This goes back and forth between the two until the image isn’t recognized as a real face by the discriminator.
The styling of these deepfakes can also be directed using instructions. You can provide prompts to suggest what you want specifically, such as: “draw me a Southeast-Asian lady with long black hair, big eyes, a sharp nose, and a symmetrical face.” Google’s Nano-Banana or DALLE-3 image creation tools would do an impressive job following this prompt. However, they can distort the eyes, the shape/size of the ears, or possibly the teeth. But with modified instructions, an individual might be able to create/edit an image that looks ultra-realistic.
According to a 2023 study by Chidera Okolie, you need basic graphic design knowledge to create deepfake images or use deepfake technology to create non-consensual pictures or videos of people, especially women. AI tools learn patterns derived from human interactions, and even mannerisms can be learned from videos of a person.
Altering Real Images With The Help of AI
A 2025 survey by the “Incogni Research Team” 1 in 4 American women have experienced some type of online abuse. The most recurrent type of abuse has been cyberbullying, and more among young females aged 18–34. Deepfake revenge porn and blackmail are the leading issues that women face because of unauthorized use of personal data or altered pictures/videos.
AI tools can take an already existing inappropriate picture off the internet and swap the face with the victim’s face. It can alter the clothes of the victim or change their pose. The background of a video or picture can easily be modified to whatever the perpetrators desire. These tools are accessible to individuals at their fingertips. There isn’t any verification needed to log into these websites, nor is there flagging when people create such inappropriate images.
Hany Farid mentions in her study how 20 years ago, people were manually making efforts to spread disinformation about political figures, yet they were very convincing. Today, people can create deepfakes without any effort and blackmail so many individuals using them. An example of a video deepfake is pupper master. In this form of deep fakeness, an individual’s(puppet master) head movements and facial expressions of one person are transferred to the other person (puppet). Unlike other face or lip-sync deep akes, a puppet-master makes it strenuous and harder to decipher the real person.
Although AI images and deepfakes have helped shape the film industry by doing the work for actors/actresses if they were asked to do any uncomfortable scenes or used for any other entertainment purposes. In the research by Chidera Okolie, she mentions that a main reason why individuals resort to image-based sexual abuse is to satisfy themselves through psychological and physical provocation of the senses. Most of these non-sexual acts are driven by the fragile masculinity of individuals, particularly the male species.
The Misuse of AI images Against Mainly Women and Other Victims
Deepfakes can affect everyone, including minors, men, trans people, politicians, or even celebrities, but predominantly women. The rising ratio of women making up to 60 to 70% of revenge porn statistics, according to a 2025 research by Eva C. Hunt. This percentage is alarming yet not surprising. The first deepfake technology was created in 2017 and posted on Reddit by an anonymous user. 2024 research by Ms. Muskan Sharma stated that the leading factors that motivate men to carry out such gruesome acts are the patriarchal society and the objectification of women. This system doesn’t hold men accountable for their actions and finds a way to point fingers at the women and blame them instead.
Non-consensual deepfakes are a major ethical issue. Data breaches, and with the extensive use of AI, people can hack into accounts, gather information, and clone individuals. Perpetrators can add movement, remove clothing, or edit images/videos to such an extent that even the victim falls for their tricks to save their dignity. These individuals blackmail the victims, saying they will expose their deepfake intimate pictures/videos to the world or just the deepfakes created from normal images. Some ex male partners use such deepfakes against their respective exes, as a form of exploitation. When they cannot handle breakups or rejection in general, they resort to such demeaning tactics.
Revenge porn is a form of revenge that men partake in to feel joy from destroying someone’s life. They love the idea of causing distress, public shame, job loss in some cases, and embarrassment to the victim. Such deepfakes can be leaked across many social media websites, as they don’t flag such content because it profits them. Some of the websites include: Reddit, Instagram, the dark web, and many porn websites. According to a survey by “Incogni Research Team,” women from mixed ethnic backgrounds have faced the most online harassment. Revenge porn, doxing, and AI deepfakes have been shown to have a severe impact on women’s lives, according to the statistics from this survey. 1 in 6 active congresswomen has been a victim of this abuse, potentially by other male competitors or followers.
Expert Source
Mashal Khan, who is currently studying law at the Chicago-Kent College of Law and has participated in research on AI ethics, said that AI deepfakes have affected so many women around her. Her own cousin was affected by a deepfake. She was blackmailed by her ex-partner, who curated inappropriate images of her without her consent. Due to having such a close relationship affected by AI, Mashal holds strong values against the rapidly growing use of AI images, and especially deepfakes.
She suggested that the tech companies should have improved software that flags an image when it is nude or includes inappropriate content. She also added that the government should approve laws in regards to safeguarding women’s rights against such deepfakes and the idea of revenge for deepfake pornography.
Mashal said, “I believe every woman has the right to feel safe enough with their partner and safe enough on the internet, rather than always questioning themselves. “What if someone uses my pictures and blackmails me?” She also adds that we need to protect the victims more than hiding the fact of who the perpetrator is.
Her approach to this inspired me to have a stronger drive towards protesting the idea that women shouldn’t always be in a constant state of panic.
Mashal also said, “I am a woman and want to feel safe from possible perpetrators, who can start blackmailing with my own pictures.” As a woman myself, I shouldn’t have to be scared of simply existing on the internet or scared of rejecting a man in case they generate deepfakes of me.
Mashal added: “Men don’t handle rejection well, and it shows, in the data presented, about the amount of revenge porn on the internet.” This concluded our conversation about this emerging issue. Women need to protect themselves against all men, in case one day they decide to become a threat to their dignity and respect.
Countermeasures Against Deepfakes
AI tools have been designed and developed by big tech companies that have the main goal of earning profit rather than following ethical practices. They do not focus on the privacy and protection of women from such perpetrators. Tackling the widespread threat of deepfakes needs a new set of laws and implications to be prevented. According to the survey from “Incogni,” there is a lack of federal data privacy laws put in place to protect people’s personal information. The legal status of AI deepfakes is unclear; existing laws regarding privacy and defamation laws do not quite cover AI-generated deepfakes.
The governments should keep up with the technological advancements, and the AI websites should have restrictions when someone tries to face-swap or should flag inappropriate images curated with it. There are laws in place for sexual abuse or pornography, but not yet for deepfake revenge pornography. These laws need a refresh and hold the awful individuals involved accountable.
So, they’re taught a lesson to never be able to mess with a woman’s reputation again. According to Chidera Okolie’s research, in Canada, deepfakes are protected under Canadian copyright laws, suggesting that they are works of artistic expression. There are laws against deepfake pornography, but only for children under the age of 18. These laws are not fair to victims older than 18 years old and need to be edited to incorporate everyone.
Another proposed solution can be that for these deepfake image generation websites, humans should be stationed to oversee what data is being flagged, and if it’s inappropriate, it should be dealt with serious consequences. Big tech companies need to focus on their users’ privacy instead of huge profit margins. If something is being created or altered with the help of AI, it should be copyrighted, and AI models should show that it’s AI-generated. There should be stricter background checks for those who can access these databases. They need a more ethical approach to the world, and awareness should be spread against such gruesome acts.
Conclusion
AI was invented to take the burden off of humans, but not 100%. The work done by AI models should be 80% and the rest of 20% should be overlooked by humans. To ensure that people are not misusing the information obtained or curated with the help of AI. AI has a lot of perks to it, as we can create images for our school projects, for websites, for movie sets, etc. But the downside to these deepfakes is copyright issues, as AI trains using real humans and their interactions. AI sparks creativity, yet some perpetrators may use their creativity for unethical means. They use deepfakes to destroy a person’s reputation, cause job losses, or even put a big strain on the victim’s personal life.
Deepfake images are a rapidly growing concern for every woman in this day and age. They are scared to post pictures online for fear of being blackmailed or doxed. Some women have started second-guessing their partners due to the rapid revenge of deepfake pornography. AI has a lot of benefits to it, but individuals are misusing its potential for their own satisfaction, and promoting the defamation and extortion of women through deepfakes.
According to my sources, women deal with more online sexual harassment and revenge porn than men. Thus, in today’s society, women are much less safe than men when it comes to blackmail and deepfakes. People should be ethically responsible when accessing such tools. Tech companies should also be more careful about who they are allowing to use such tools.
—–
FAQ
Q: What are AI deepfakes, and how are they created?
AI deepfakes are images or videos created using Artificial Intelligence technologies, often resulting in non-consensual exploitation. AI tools are actively training on and have been trained using real human faces and interactions, sometimes using real images stolen by their respective owners. The creation process often involves a generative adversarial network (GAN). A GAN is adversarial because it has two parts: 1. A generator that makes the fake images. 2. A discriminator that determines if the image produced is real or fake. This process goes back and forth until the image is not recognized as a real face by the discriminator.
Q: Why is it becoming harder to distinguish AI deepfakes from real images?
At first, it was easy to distinguish an AI image from a real one. However, as AI models are training every day, 24/7, to become better at their jobs, it has become harder to note the difference sometimes. The realistic nature of these images is becoming the new norm.
Q: Besides creating new images, how are AI tools used to alter existing images?
AI tools can take an already existing inappropriate picture off the internet and swap the face with the victim’s face. They can also alter the victim’s clothes or change their pose. The background of a video or picture can easily be modified to whatever the perpetrators desire, misuse, victims, and consequences.
Q: How are AI deepfakes primarily being misused, and who are the main targets?
AI deepfakes are being misused for exploitation, especially through blackmail and deepfake revenge porn. They are used in a non-consensual manner to extort and defame women. While deepfakes can affect various groups — including minors, men, trans people, politicians, and celebrities — they predominantly target women. Women make up between 60% to 70% of revenge porn statistics, according to a 2025 research study.
Q: Are children also victims of deepfake misuse?
Yes, individuals have taken AI images/videos of children being sexually abused and altered them to put on the dark web, which is described as horrific.
Q: What motivates perpetrators to create non-consensual deepfakes, particularly against women?
A main reason individuals resort to image-based sexual abuse is to satisfy themselves through the psychological and physical provocation of the senses. Most of these non-sexual acts are driven by the fragile masculinity of individuals, particularly males. Leading factors include the patriarchal society and the objectification of women. Furthermore, ex-male partners use such deepfakes as a form of exploitation when they cannot handle breakups or rejection in general. Revenge porn is seen as a way to feel joy from destroying someone’s life, causing distress, public shame, job loss, and embarrassment to the victim.
Q: What are the consequences of deepfake misuse on victims?
Deepfakes can destroy a person’s reputation, cause job losses, or put a significant strain on the victim’s personal life. Victims may be blackmailed by perpetrators who threaten to expose intimate deepfake pictures or videos. Deepfake images are causing women to be scared to post pictures online for fear of being blackmailed or doxed, and some women have started second-guessing their partners.
Q: What is the current legal status of AI deepfakes?
The legal status of AI deepfakes is often unclear. Existing laws regarding privacy and defamation do not quite cover AI-generated deepfakes. Currently, according to one research study, deepfakes in Canada are protected under Canadian copyright laws, suggesting they are works of artistic expression. There are laws against deepfake pornography, but only for children under the age of 18, which is deemed unfair to victims older than 18.
Q: What countermeasures and legislative changes are suggested to combat deepfake misuse?
A new set of laws and implications is needed to tackle the widespread threat of deepfakes. Proposed solutions include: Governments should approve laws specifically safeguarding women’s rights against such deepfakes and the idea of revenge for deepfake pornography, the laws for sexual abuse and pornography need a refresh to incorporate deepfake revenge pornography, and hold awful individuals accountable. Tech companies should implement improved software that flags an image when it is nude or includes inappropriate content.