There is a crisis unfolding online that most people never see. Millions of women and girls wake up each day unaware that AI-generated sexual images of them are being created, shared, and monetized without their knowledge or consent. What began as a niche technological curiosity in 2017 has become a global industry of harm, one powered by the same artificial intelligence tools being celebrated for transforming business and daily life.
This is not a hypothetical threat. It is happening right now, at scale, and the numbers are staggering.
The word "deepfake" was born in infamy. In late 2017, a Reddit user operating under the handle "deepfakes" created a subreddit where he and others shared pornographic videos featuring the faces of celebrities, including Gal Gadot, Scarlett Johansson, and Taylor Swift, digitally grafted onto the bodies of adult performers. The community grew to over 15,000 followers before Reddit banned the subreddit in February 2018 for violating its "involuntary pornography" policies.
The ban was too little, too late. By then, the technological genie was out of the bottle.
In January 2018, a freely available desktop tool called FakeApp was released to the public. Built on the same deep learning principles used by major tech companies, FakeApp required no coding knowledge and could run on consumer hardware. Shortly after, tools like FaceSwap and DeepFaceLab lowered the barrier further. Suddenly, anyone with a laptop and a photo of a real person could fabricate explicit imagery. A 2019 report by Dutch cybersecurity startup Deeptrace found that 96% of all deepfake videos online were pornographic , a statistic that has held remarkably steady in the years since.
The platform bans and public outrage of 2018 did not stop the technology. They merely pushed it underground, where it matured rapidly.
The deepfake problem of 2018 and the deepfake problem of 2026 are not the same problem. The difference is generative AI.
The arrival of large-scale image synthesis models, including Stable Diffusion, Midjourney, and dozens of purpose-built applications, transformed deepfake pornography from a technically demanding craft into a point-and-click operation. Websites branded as "nudification" or "undress" apps began appearing across the internet, promising users the ability to upload a photo of any woman and receive a fabricated nude image in seconds. In August 2024, the San Francisco City Attorney's office sued 16 such websites in a landmark legal action.
The scale of the resulting harm is almost incomprehensible.
According to multiple sources tracking deepfake content, the volume of deepfake videos online grew from approximately 14,000 in 2019 to a projected 8 million by 2025, a 571-fold increase in six years. The volume of deepfake pornography produced in 2023 alone was 464% higher than in 2022, reflecting the exponential acceleration that generative AI enabled. Researchers consistently find that 96–98% of all deepfake content is non-consensual pornography, and an estimated 99% of victims are women and girls.
The crisis has moved from celebrity targets to ordinary people. When you can upload a photo from Instagram and have an AI generate explicit imagery in under a minute, the potential victim pool becomes everyone.
Understanding this crisis requires understanding how it works technically, because the sophistication is precisely what makes it so dangerous.
The original deepfake method involves training a neural network on images of a target person's face, then using that network to map the face onto a different person's body in video or imagery. Early versions required hundreds of source photos and hours of computing time. Modern tools can accomplish the same task with a single image and complete the process in seconds, using cloud-based infrastructure that requires no local hardware at all.
Newer generative AI systems don't even require a face swap. Tools built on diffusion model architectures can take a fully-clothed image of a real person and generate a fabricated nude — maintaining enough physical similarity to be recognizable as the target individual. These "undress apps" became one of the fastest-growing categories of abusive AI applications. In December 2025, xAI, the company owned by Elon Musk behind the Grok AI assistant integrated into the social platform X, updated Grok to allow users to upload images and request edits, a capability that was immediately exploited to create non-consensual explicit content.
Alibaba's 2025 AI video generation model Wan 2.1 was similarly modified by bad actors to produce non-consensual pornographic video.
Once created, deepfake pornography spreads through a well-established pipeline. Research suggests that 55% of deepfake pornography spreads via Telegram, with nearly half first appearing on X (Twitter) before removal. Dedicated deepfake pornography websites, many hosted in jurisdictions with permissive or nonexistent laws, host and monetize the content. Some creators charge subscription fees for access to deepfakes of specific targets, creating a de facto market in other people's violated identities.
Human beings are alarmingly bad at detecting AI-generated imagery: studies show people correctly identify high-quality deepfake video only **24.5% of the time**, barely better than a coin flip.
In late January 2024, explicit AI-generated images of pop star Taylor Swift were mass-distributed on 4chan and X. One post alone was viewed 47 million times before its removal. The incident was not just a violation of a single famous woman, it was a breaking-point moment that made the deepfake pornography crisis impossible for policymakers to ignore.
The current White House press secretary, Karine Jean-Pierre, called the images "alarming." Representative Joe Morelle of New York, who had been trying to advance federal legislation for years, described renewed urgency in the wake of the incident. The bipartisan No AI Fraud Act, introduced in the House, would give victims civil legal recourse against creators and distributors of non-consensual AI imagery.
If the Taylor Swift incident illustrated the celebrity dimension of the crisis, South Korea offered a preview of what the crisis looks like when it reaches ordinary people at scale. In mid-2024, South Korean authorities uncovered extensive networks of AI deepfake pornography Telegram chatrooms that were targeting schools and universities. Teenage boys were using AI to fabricate explicit images of female classmates and teachers from innocent Instagram photos.
The numbers were sobering: between 2021 and July 2024, 793 deepfake sex crimes were reported in South Korea but only 16 people were arrested. After the Telegram chatroom story broke publicly in August 2024, 118 new cases were reported in just five days. The crisis drew international attention because it illustrated that this technology had moved beyond targeting celebrities, it now threatened any woman or girl whose photograph existed online.
The most urgent dimension of the crisis involves children. The National Center for Missing and Exploited Children (NCMEC) reported that AI-related child sexual exploitation reports rose from 6,835 in the first half of 2024 to 440,419 in the first half of 2025, a staggering 6,443% increase in a single year. As of August 2025, only 45 U.S. states had enacted laws explicitly criminalizing AI-generated child sexual abuse material, leaving significant legal gaps. Even purely synthetic AI-generated child sexual abuse material, involving no identifiable real child, is harmful because it normalizes exploitation and diverts law enforcement resources.
For years, legal frameworks failed to keep pace with the technology. Many jurisdictions had laws against non-consensual pornography involving real photographs but nothing explicitly addressing AI-fabricated imagery. Perpetrators exploited these gaps deliberately.
That changed in May 2025. The TAKE IT DOWN Act (Tools to Address Known Exploitation by Immobilizing Technological Deepfakes on Websites and Networks) was signed into law by President Donald Trump on May 19, 2025, representing the first major federal legislation directly addressing AI-generated sexual exploitation.12 The bill passed both chambers of Congress by near-unanimous margins.
Key provisions of the law include:
The law moved quickly to its first use. In April 2026, James Strahler II, a 37-year-old Ohio man, became the first person convicted under the TAKE IT DOWN Act. Strahler had used AI to create non-consensual intimate imagery of both adult and minor victims in his neighborhood and shared the material on child exploitation websites. He pleaded guilty to cyberstalking, producing obscene child sexual abuse imagery, and publishing digital forgeries.
Months later, in May 2026, federal prosecutors charged Cornelius Shannon, 51, and Arturo Hernandez, 20, under the same law. Shannon had published at least 360 albums of AI-generated pornography depicting approximately 90 women, including politicians and celebrities, accumulating nearly 3 million views. Together, the two defendants were charged with crimes involving 140 named victims.
Federal action has been accompanied by a wave of state legislation. 46 states have now enacted laws targeting AI-generated synthetic media. Tennessee's ELVIS Act (Ensuring Likeness, Voice, and Image Security Act), enacted in 2024, specifically prohibits AI replication of a person's voice and likeness without consent, a model that other states have begun to adopt.14
Internationally, France criminalized non-consensual sexual deepfakes through Article 226-8-1 of the Penal Code in 2024. Proposed UK amendments would impose up to two years imprisonment for creating sexually explicit deepfakes with intent to cause distress. South Korea, galvanized by its 2024 school crisis, significantly strengthened its criminal penalties for deepfake sex crimes.
Still, the global patchwork of laws remains uneven, and perpetrators continue to exploit jurisdictional gaps.
The deepfake pornography industry has emerged precisely because it could. The technology became widely available. The legal protections didn't exist. The platforms were slow to act. And the people most harmed, women, girls, and children, were the ones least likely to have the resources or political standing to fight back.
What has changed is momentum. The TAKE IT DOWN Act establishes that fabricated sexual imagery is not a "victimless" digital curiosity - it is a crime. The first prosecutions under that law send a signal that will compound as enforcement expands. Detection technology is improving; the deepfake detection market is projected to reach **$15.7 billion by 2026**, growing at roughly 42% annually.
This has an impact on the financial markets, especially values-based investors. Inspire has directly encountered this issue with the launch of SpaceX (SPCX), which emerged as one of the largest IPO debuts in history, initially valued at $75 billion (approximately 555.5 million shares at $135 per share). While SpaceX’s primary mission centers on aerospace innovation and exploration, deepfake pornography and broader pornographic content have emerged as reputational concerns within its corporate ecosystem. As referenced above, xAI, a subsidiary of SpaceX and the developer of the Grok AI assistant integrated into X, updated Grok to allow users to upload images and request edits, a capability that was immediately exploited to generate non-consensual explicit content at scale. The resulting volume of deepfake imagery prompted significant restrictions from xAI, accelerated in part by threatened regulatory action from the EU, which threatened to ban Grok and potentially even bring legal action against Musk. Despite those restrictions, Grok retains image-to-image generation capabilities that can be exploited to produce graphic content, along with features supporting nudity and sexually explicit material in text and voice formats. These factors contributed to SpaceX receiving a score of -81 on Inspire Insight, upon launch.
But law and technology alone will not solve this. The companies building generative AI tools must not turn a blind eye to the possible abuse these models are capable of. Platforms that host this content have a responsibility to remove it swiftly and report creators. And all of us, as consumers of technology and participants in the digital economy, have a role in demanding accountability from the companies we support.
The same AI tools enabling remarkable breakthroughs in medicine, education, and productivity are being used to strip human dignity from millions of people who never consented to any of it. That tension demands a response, not only from lawmakers, but from anyone who believes that how we build technology reflects what we value as a society.
This is not merely a technology story. It is a story about man’s sin nature, human dignity, and the weaponization of innovation against people who have no idea they are being violated.
The rapid expansion of AI capabilities carries a familiar tension: powerful tools, placed in the hands of those governed by unchecked sinful desire, become instruments of harm. While AI tools in general cannot be labeled as “bad” or “sinful” on their own, they offer the capacity for people with dark motivations and minimal self-control to hurt people through the anonymity of the digital wild west. The pattern is consistent across the history of pornography. Print, video, the internet, and social media have each been exploited for sexual gratification nearly as soon as they were created. At the root of this pattern lies the same cause: man’s sinful nature, apart from the sanctifying work of Christ Jesus and the transforming authority of the Holy Spirit to overcome our flesh (Colossians 3:5; Galatians 5:24). Legislative action, corporate accountability, and platform policy are essential responses, however imperfect they might be, and are worthy of support. But for the investor who follows Christ, and for every believer, the most powerful response remains vigilant awareness, responsible stewardship, and, most importantly, prayer - prayer that God’s light would pierce these dark places, and that His justice and peace would prevail on earth as in heaven.
“For this is the will of God, your sanctification: that you abstain from sexual immorality; that each one of you know how to control his own body in holiness and honor, not in the passion of lust like the Gentiles who do not know God;” 1 Thessalonians 4:3-5
Bringing ten years of experience as a professor at a Christian college, Chris is deeply rooted in faith-based values. He excels as a faith-based writer, adept at crafting financial curriculum with spiritual depth. In his role, Chris is responsible for sourcing and evaluating reliable sources to accurately screen companies and assign them an Inspire Impact Score. Furthermore, he meticulously curates screening categories to maintain clarity and alignment with Inspire's faith-based mission.
All expressions of opinion are subject to change without notice. Investing involves risk, including the possible loss of principal.
There is a crisis unfolding online that most people never see. Millions of women and girls wake up each day unaware that AI-generated sexual images of them are being created, shared, and monetized without their knowledge or consent. What began as a niche technological curiosity in 2017 has become a global industry of harm, one powered by the same artificial intelligence tools being celebrated for transforming business and daily life.
This is not a hypothetical threat. It is happening right now, at scale, and the numbers are staggering.
The word "deepfake" was born in infamy. In late 2017, a Reddit user operating under the handle "deepfakes" created a subreddit where he and others shared pornographic videos featuring the faces of celebrities, including Gal Gadot, Scarlett Johansson, and Taylor Swift, digitally grafted onto the bodies of adult performers. The community grew to over 15,000 followers before Reddit banned the subreddit in February 2018 for violating its "involuntary pornography" policies.
The ban was too little, too late. By then, the technological genie was out of the bottle.
In January 2018, a freely available desktop tool called FakeApp was released to the public. Built on the same deep learning principles used by major tech companies, FakeApp required no coding knowledge and could run on consumer hardware. Shortly after, tools like FaceSwap and DeepFaceLab lowered the barrier further. Suddenly, anyone with a laptop and a photo of a real person could fabricate explicit imagery. A 2019 report by Dutch cybersecurity startup Deeptrace found that 96% of all deepfake videos online were pornographic , a statistic that has held remarkably steady in the years since.
The platform bans and public outrage of 2018 did not stop the technology. They merely pushed it underground, where it matured rapidly.
The deepfake problem of 2018 and the deepfake problem of 2026 are not the same problem. The difference is generative AI.
The arrival of large-scale image synthesis models, including Stable Diffusion, Midjourney, and dozens of purpose-built applications, transformed deepfake pornography from a technically demanding craft into a point-and-click operation. Websites branded as "nudification" or "undress" apps began appearing across the internet, promising users the ability to upload a photo of any woman and receive a fabricated nude image in seconds. In August 2024, the San Francisco City Attorney's office sued 16 such websites in a landmark legal action.
The scale of the resulting harm is almost incomprehensible.
According to multiple sources tracking deepfake content, the volume of deepfake videos online grew from approximately 14,000 in 2019 to a projected 8 million by 2025, a 571-fold increase in six years. The volume of deepfake pornography produced in 2023 alone was 464% higher than in 2022, reflecting the exponential acceleration that generative AI enabled. Researchers consistently find that 96–98% of all deepfake content is non-consensual pornography, and an estimated 99% of victims are women and girls.
The crisis has moved from celebrity targets to ordinary people. When you can upload a photo from Instagram and have an AI generate explicit imagery in under a minute, the potential victim pool becomes everyone.
Understanding this crisis requires understanding how it works technically, because the sophistication is precisely what makes it so dangerous.
The original deepfake method involves training a neural network on images of a target person's face, then using that network to map the face onto a different person's body in video or imagery. Early versions required hundreds of source photos and hours of computing time. Modern tools can accomplish the same task with a single image and complete the process in seconds, using cloud-based infrastructure that requires no local hardware at all.
Newer generative AI systems don't even require a face swap. Tools built on diffusion model architectures can take a fully-clothed image of a real person and generate a fabricated nude — maintaining enough physical similarity to be recognizable as the target individual. These "undress apps" became one of the fastest-growing categories of abusive AI applications. In December 2025, xAI, the company owned by Elon Musk behind the Grok AI assistant integrated into the social platform X, updated Grok to allow users to upload images and request edits, a capability that was immediately exploited to create non-consensual explicit content.
Alibaba's 2025 AI video generation model Wan 2.1 was similarly modified by bad actors to produce non-consensual pornographic video.
Once created, deepfake pornography spreads through a well-established pipeline. Research suggests that 55% of deepfake pornography spreads via Telegram, with nearly half first appearing on X (Twitter) before removal. Dedicated deepfake pornography websites, many hosted in jurisdictions with permissive or nonexistent laws, host and monetize the content. Some creators charge subscription fees for access to deepfakes of specific targets, creating a de facto market in other people's violated identities.
Human beings are alarmingly bad at detecting AI-generated imagery: studies show people correctly identify high-quality deepfake video only **24.5% of the time**, barely better than a coin flip.
In late January 2024, explicit AI-generated images of pop star Taylor Swift were mass-distributed on 4chan and X. One post alone was viewed 47 million times before its removal. The incident was not just a violation of a single famous woman, it was a breaking-point moment that made the deepfake pornography crisis impossible for policymakers to ignore.
The current White House press secretary, Karine Jean-Pierre, called the images "alarming." Representative Joe Morelle of New York, who had been trying to advance federal legislation for years, described renewed urgency in the wake of the incident. The bipartisan No AI Fraud Act, introduced in the House, would give victims civil legal recourse against creators and distributors of non-consensual AI imagery.
If the Taylor Swift incident illustrated the celebrity dimension of the crisis, South Korea offered a preview of what the crisis looks like when it reaches ordinary people at scale. In mid-2024, South Korean authorities uncovered extensive networks of AI deepfake pornography Telegram chatrooms that were targeting schools and universities. Teenage boys were using AI to fabricate explicit images of female classmates and teachers from innocent Instagram photos.
The numbers were sobering: between 2021 and July 2024, 793 deepfake sex crimes were reported in South Korea but only 16 people were arrested. After the Telegram chatroom story broke publicly in August 2024, 118 new cases were reported in just five days. The crisis drew international attention because it illustrated that this technology had moved beyond targeting celebrities, it now threatened any woman or girl whose photograph existed online.
The most urgent dimension of the crisis involves children. The National Center for Missing and Exploited Children (NCMEC) reported that AI-related child sexual exploitation reports rose from 6,835 in the first half of 2024 to 440,419 in the first half of 2025, a staggering 6,443% increase in a single year. As of August 2025, only 45 U.S. states had enacted laws explicitly criminalizing AI-generated child sexual abuse material, leaving significant legal gaps. Even purely synthetic AI-generated child sexual abuse material, involving no identifiable real child, is harmful because it normalizes exploitation and diverts law enforcement resources.
For years, legal frameworks failed to keep pace with the technology. Many jurisdictions had laws against non-consensual pornography involving real photographs but nothing explicitly addressing AI-fabricated imagery. Perpetrators exploited these gaps deliberately.
That changed in May 2025. The TAKE IT DOWN Act (Tools to Address Known Exploitation by Immobilizing Technological Deepfakes on Websites and Networks) was signed into law by President Donald Trump on May 19, 2025, representing the first major federal legislation directly addressing AI-generated sexual exploitation.12 The bill passed both chambers of Congress by near-unanimous margins.
Key provisions of the law include:
The law moved quickly to its first use. In April 2026, James Strahler II, a 37-year-old Ohio man, became the first person convicted under the TAKE IT DOWN Act. Strahler had used AI to create non-consensual intimate imagery of both adult and minor victims in his neighborhood and shared the material on child exploitation websites. He pleaded guilty to cyberstalking, producing obscene child sexual abuse imagery, and publishing digital forgeries.
Months later, in May 2026, federal prosecutors charged Cornelius Shannon, 51, and Arturo Hernandez, 20, under the same law. Shannon had published at least 360 albums of AI-generated pornography depicting approximately 90 women, including politicians and celebrities, accumulating nearly 3 million views. Together, the two defendants were charged with crimes involving 140 named victims.
Federal action has been accompanied by a wave of state legislation. 46 states have now enacted laws targeting AI-generated synthetic media. Tennessee's ELVIS Act (Ensuring Likeness, Voice, and Image Security Act), enacted in 2024, specifically prohibits AI replication of a person's voice and likeness without consent, a model that other states have begun to adopt.14
Internationally, France criminalized non-consensual sexual deepfakes through Article 226-8-1 of the Penal Code in 2024. Proposed UK amendments would impose up to two years imprisonment for creating sexually explicit deepfakes with intent to cause distress. South Korea, galvanized by its 2024 school crisis, significantly strengthened its criminal penalties for deepfake sex crimes.
Still, the global patchwork of laws remains uneven, and perpetrators continue to exploit jurisdictional gaps.
The deepfake pornography industry has emerged precisely because it could. The technology became widely available. The legal protections didn't exist. The platforms were slow to act. And the people most harmed, women, girls, and children, were the ones least likely to have the resources or political standing to fight back.
What has changed is momentum. The TAKE IT DOWN Act establishes that fabricated sexual imagery is not a "victimless" digital curiosity - it is a crime. The first prosecutions under that law send a signal that will compound as enforcement expands. Detection technology is improving; the deepfake detection market is projected to reach **$15.7 billion by 2026**, growing at roughly 42% annually.
This has an impact on the financial markets, especially values-based investors. Inspire has directly encountered this issue with the launch of SpaceX (SPCX), which emerged as one of the largest IPO debuts in history, initially valued at $75 billion (approximately 555.5 million shares at $135 per share). While SpaceX’s primary mission centers on aerospace innovation and exploration, deepfake pornography and broader pornographic content have emerged as reputational concerns within its corporate ecosystem. As referenced above, xAI, a subsidiary of SpaceX and the developer of the Grok AI assistant integrated into X, updated Grok to allow users to upload images and request edits, a capability that was immediately exploited to generate non-consensual explicit content at scale. The resulting volume of deepfake imagery prompted significant restrictions from xAI, accelerated in part by threatened regulatory action from the EU, which threatened to ban Grok and potentially even bring legal action against Musk. Despite those restrictions, Grok retains image-to-image generation capabilities that can be exploited to produce graphic content, along with features supporting nudity and sexually explicit material in text and voice formats. These factors contributed to SpaceX receiving a score of -81 on Inspire Insight, upon launch.
But law and technology alone will not solve this. The companies building generative AI tools must not turn a blind eye to the possible abuse these models are capable of. Platforms that host this content have a responsibility to remove it swiftly and report creators. And all of us, as consumers of technology and participants in the digital economy, have a role in demanding accountability from the companies we support.
The same AI tools enabling remarkable breakthroughs in medicine, education, and productivity are being used to strip human dignity from millions of people who never consented to any of it. That tension demands a response, not only from lawmakers, but from anyone who believes that how we build technology reflects what we value as a society.
This is not merely a technology story. It is a story about man’s sin nature, human dignity, and the weaponization of innovation against people who have no idea they are being violated.
The rapid expansion of AI capabilities carries a familiar tension: powerful tools, placed in the hands of those governed by unchecked sinful desire, become instruments of harm. While AI tools in general cannot be labeled as “bad” or “sinful” on their own, they offer the capacity for people with dark motivations and minimal self-control to hurt people through the anonymity of the digital wild west. The pattern is consistent across the history of pornography. Print, video, the internet, and social media have each been exploited for sexual gratification nearly as soon as they were created. At the root of this pattern lies the same cause: man’s sinful nature, apart from the sanctifying work of Christ Jesus and the transforming authority of the Holy Spirit to overcome our flesh (Colossians 3:5; Galatians 5:24). Legislative action, corporate accountability, and platform policy are essential responses, however imperfect they might be, and are worthy of support. But for the investor who follows Christ, and for every believer, the most powerful response remains vigilant awareness, responsible stewardship, and, most importantly, prayer - prayer that God’s light would pierce these dark places, and that His justice and peace would prevail on earth as in heaven.
“For this is the will of God, your sanctification: that you abstain from sexual immorality; that each one of you know how to control his own body in holiness and honor, not in the passion of lust like the Gentiles who do not know God;” 1 Thessalonians 4:3-5
