South Korea's Deepfake Pornography Crisis Expands to Schools and Workplaces
Hundreds of young men and teenage boys in South Korea used AI-based deepfake technology to create explicit synthetic images and videos of women, girls, teachers, and family members, distributing the material through encrypted chat rooms with up to 220,000 members. Police arrested dozens of perpetrators while acknowledging the technology's sophistication makes detection difficult, and observers noted the government's initial response was slower than in a similar 2019 case.
What happened
In August 2024, a surge of deepfake pornography created using AI technology was exposed in South Korea, involving hundreds of young men and teenage boys who created explicit images and videos of classmates, teachers, military colleagues, and other women and girls, including minors, by taking their social media photos and using deepfake apps to attach their faces to pornographic bodies.
The material was spread through encrypted messaging service Telegram chat rooms, some with as many as 220,000 members. Investigators noted the technology is sophisticated enough that ordinary people often cannot distinguish the deepfakes from real images.
From January to July 2024, South Korean authorities arrested 178 suspected perpetrators related to deepfake sexual exploitation crimes, of which 131 were teenagers. By mid-August, approximately 15 people had been arrested, with some perpetrators fleeing overseas.
Reported impact
Women and girls reported severe psychological distress, with some experiencing anxiety about future implications and uncertainty about how widely their images had been distributed. Teachers found themselves depicted in explicit deepfake material by their own students.
The abuse extended to family members, with some chat rooms specifically dedicated to creating deepfakes of victims' mothers, sisters, and other relatives. Victims expressed that they feared these images would follow them for life.
Index analysis
According to SecurityHero, South Korea accounted for 53% of global deepfake sexual content in 2023, despite the country representing less than 2% of global GDP. Volunteers documented perpetrator locations on a map showing deepfake-related crimes concentrated across schools, universities, hospitals, and military installations throughout South Korea's central regions.
The incident is described as similar to the 2019 Nๅทๆฟ case but with lower barriers to entry due to AI technology reducing the cost and effort needed to produce exploitative content. However, public and governmental response was reported as more muted compared to the earlier scandal.
Governance controls
The National Assembly had enacted a vaguely worded law against making and distributing deepfake pornography following an earlier 2020 blackmail case, but this did not prevent the crime wave from occurring.
Some observers noted that AI-generated content was initially not prioritized by police and government authorities because it was perceived as not real, creating a gap in enforcement and deterrence.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.