A Death in Cryptoland Introduces: Deepfake Porn Empire

February 24, 2026·39 min·3 clips

Twitch streamer CutieCinderella wakes up to find deepfake porn of herself viral on Twitter, sparking a personal crisis.

The episode opens in media res with Twitch streamer CutieCinderella describing waking up to find herself trending on Twitter after a colleague named Atriok accidentally revealed a deepfake porn site during a live broadcast. Sam Cole, the host and a tech journalist at 404 Media, frames this moment as the mainstream breaking point for an issue she has been reporting on since 2017. Cole traces the history of fake images back to 1840 when French photographer Hippolyte Bayard faked his own death in a photograph, and to the Cottingley Fairies hoax of 1917 that fooled Arthur Conan Doyle. Digital fakery accelerated in the 1990s with Photoshop, which quickly attracted porn creators, and a Canadian libertarian named Cary Pearson, known as Lux Luker, ran two Usenet groups and a website dedicated to celebrity nude faking from 1996. A pseudonymous participant called TMFU describes spending every night producing and sharing Photoshop fakes in Luker's community, which had rules nominally prohibiting fakes of minors but published a chart of underage celebrities with the years they would turn 18. Hollywood's pioneering use of CGI in Jurassic Park and the digital recreation of Paul Walker's face for Fast and Furious 7 tested public tolerance for photorealistic AI imagery. The decisive technological breakthrough came in 2014 when Ian Goodfellow, a Montreal PhD student, invented Generative Adversarial Networks in a single night after a bar argument, then published and open-sourced the code. The system works by pitting two algorithms against each other: one generates fake images, the other tries to detect fakes, and they improve by competing. By December 2017, a Reddit user named DeepFakes had used this technology to produce fake porn videos featuring recognizable celebrity faces and posted them publicly. Sam Cole broke the story at Motherboard, noting the DeepFakes creator saw no difference between his work and Hollywood CGI. Despite 96% of deepfake videos being non-consensual porn according to multiple research studies, mainstream media and government attention focused on political deepfakes, with DARPA spending tens of millions on detection technology. The political deepfakes feared most — of world leaders launching wars — were largely debunked quickly in practice. The Cutie Cinderella incident in January 2023 finally brought deepfake porn to mass public attention and put a human face on the harm. Cutie describes the psychological effects including flashbacks to childhood sexual assault, family members seeing the images, and knowing the person exposed watching was her boyfriend's best friend. Atriok apologized on stream, donated $60,000 to a takedown law firm, and the deepfake site took the videos down. The season promises an investigative trail across three countries and four newsrooms leading to the operator of the world's largest non-consensual deepfake site.

As heard by us

A sober account of how deepfake porn developed, and why the harm spread so quickly.

The episode opens with CutieCinderella's shock at waking up to Twitter attention, then moves into the machinery behind deepfake porn. It traces the shift from fake celebrity porn assembled by hand to newer automated image-making methods, with Ian Goodfellow and generative models…

Read the full review in PlayNext →

Why you'd press play

Watch one viral scare turn into a clear history of fake porn and generative AI.

Read the full recommendation in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from A Death in Cryptoland