South Korea passes bill targeting deepfake sexual exploitation

A South Korean parliamentary committee has approved a bill that would impose prison sentences for possessing or viewing deepfake sexual content, Yonhap news agency reports. The bill comes in response to growing public concern about digital sex crimes, particularly after the discovery of Telegram chatrooms where AI-generated deepfake pornography was being shared with female students and staff.

The revised law imposes stiff penalties on those found to possess, purchase, store or view deepfake sexual content, with possible prison sentences of up to three years or fines of up to 30 million won (about $22,500). Lawmakers from both the ruling and opposition parties included a provision to exempt those who “unknowingly” come into contact with such materials from punishment.

The committee also made revisions to the Protection of Children from Sexual Offences Act, introducing tougher penalties for those who use exploitative materials to blackmail or coerce minors. Under the new provisions, sentences for blackmail are increased to a minimum of three years, while coercion can result in sentences of five years or more.

The revisions also concern the Sexual Violence Prevention Act, which emphasises the government’s responsibility to remove illegally recorded content and to help victims reintegrate into society.

The increased efforts to combat sexual deepfakes coincide with the Pavel Durov case. Telegram founder Durov is currently under formal investigation in France as authorities probe organized crime activity linked to the messaging platform.

Deepfake technology refers to a type of artificial intelligence used to create convincing fake images, videos, and audio recordings. This technology can manipulate existing content, swapping one person for another, or generating entirely new scenarios in which individuals appear to say or do things they never actually did. The greatest danger of deepfakes lies in their ability to spread false information that appears to come from trusted sources.

“I’ve always been cautious about giving out personal information or photos,” said Ines Kwon, a university student. “But with so many crimes these days involving AI and deepfake technology, I find myself being even more cautious.”

Last week, South Korea’s National Police Agency announced plans to invest 2.7 billion won ($2 million) annually over the next three years to develop deep-learning technology to detect fabricated content, including deepfakes and voice cloning. The agency will also upgrade its current software for monitoring AI-generated videos.

The urgency of these measures is underscored by alarming statistics: the number of cases of cyber sexual violence this year is 11 times higher than in 2018, yet there have been only 16 arrests in 793 reported deepfake crimes since 2021, police data shows.

“It’s been six years since I fell victim to AI-generated deepfake pornography. As I transitioned from being a student to a working professional, it would be a lie to say the incident didn’t leave a mark on me,” said one anonymous victim of a deepfake crime.

“But when I think of other women who could be experiencing the same or even worse suffering, it ignites a fire in me to become stronger and stand up to this.”

President Yoon Suk-yeol said deepfakes constitute a serious crime that undermines social harmony and urged relevant ministries to take decisive action. In a joint statement in late August, 84 women’s organizations stressed that the root cause of the deepfake crisis is “structural gender discrimination” and called for gender equality as a fundamental solution.

Oh Kyung-Jin, secretary general of the Korean Women’s Association, said the fundamental problem is the “culture of misogyny,” where women are seen only as sex objects, targets of crime and forms of entertainment, rather than as equal citizens, especially among teenagers.