Digital Sexual Crime Materials Automatically Detected and Reported 24/7 by AI... "Processing Time per Case: 1 Minute"

Ministry of Gender Equality and Family Automates Deletion Requests
AI-Based Exploitation Detection and Deepfake Identification System Introduced

The process of deleting digital sexual crime victim videos will be changed to an automated processing system. Previously, staff members had to manually search for and delete each video, but with the introduction of artificial intelligence (AI), the processing time per case is expected to be reduced to less than one minute.

On April 1, the Ministry of Gender Equality and Family announced that it would automate the entire process for approximately 20,000 websites, from "requests for deletion of victim videos" to "management of processing history."

Officials at the Digital Sexual Crime Victim Support Center of the Korea Women’s Human Rights Institute are deleting illegal videos. Photo by Jo Yongjun

The Central Digital Crime Victim Support Center (Digital Crime Victim Center) has linked its deletion support system with the U.S. National Center for Missing & Exploited Children (NCMEC) and the global Content Delivery Network (CDN) operator Cloudflare, making it possible to automatically request mass deletion of URLs that allow bypass access.

The Google content removal form and the Telecommunications Business Act reporting form have also been improved to allow for automatic completion and submission. In addition, a filter function (grayscale processing) for recorded materials has been introduced to help prevent emotional exhaustion among staff members.

The Ministry of Gender Equality and Family stated, "With the adoption of this system, we have addressed the limitations of the previous manual and reactive approach to victim protection, and by automating and enhancing the intelligence of the digital sexual crime response system, we have significantly improved the speed and accuracy of our response."

AI technology will also be used to automatically collect and request deletion of child and youth sexual exploitation materials or luring information exposed on social networking services (SNS) and random chat apps, operating 24 hours a day.

The AI model implemented in the system determines whether content constitutes child or youth sexual exploitation material by analyzing ▲keyword detection, ▲text extraction from images, ▲assessment of nudity, and ▲comparison of similar images.

Results from the pilot operation showed that the average daily number of cases collected per staff member increased dramatically, with child and youth sexual exploitation materials rising by more than 2.7 times and sexual exploitation luring information increasing by more than 80 times.

The government plans to support the prompt deletion of child and youth sexual exploitation materials through the full-scale operation of the system. For sexual exploitation luring information, the government will operate "online outreach," where a professional counselor is involved alongside the reporting process.

With the introduction of the "AI-based deepfake detection solution," it is now possible to identify synthetic content that was previously difficult to detect with the naked eye, enabling video deletion before further distribution. Furthermore, if videos suspected of being synthetic or edited are found during online monitoring, the deepfake detection solution can identify and delete synthetic content even if the victim is unaware, helping to alleviate victims' anxiety.

Minister of Gender Equality and Family Won Min-kyung said, "By introducing AI technology in the field, we will improve the speed and accuracy of our response and build a comprehensive support system to take the lead in helping victims recover their daily lives quickly."

This content was produced with the assistance of AI translation services.