From 3 hours to 6 minutes: How Seoul’s new AI shield protects victims of cyber crimes

The Seoul Metropolitan Government announced on Wednesday the nationwide rollout of a proprietary, 24-hour artificial intelligence system designed to detect and eliminate digital sex crimes in near real-time.

The technology, which has undergone three years of iterative development, is being offered free of charge to domestic and international institutions to combat the rapid proliferation of deepfakes and non-consensual illicit content.

A Quantum Leap in Speed

City officials stated that the AI-integrated system has fundamentally altered the timeline of victim support. While human counsellors previously spent approximately three hours manually searching and reporting a single piece of illegal content, the AI completes the cycle, from detection to reporting, in just six minutes.

This 30-fold increase in speed is critical in the digital age, where images can be mirrored across thousands of servers within hours. By drastically shortening the "window of exposure," the city aims to mitigate the long-term psychological trauma suffered by victims.

Beyond Image Matching: The Tech Behind the Shield

Unlike traditional "hash-matching" software, which requires an exact digital fingerprint of an original file to find copies, Seoul’s new AI utilises multi-modal deep learning.

• Behavioural Analysis: The system identifies illicit material by analysing a combination of video patterns, audio cues, and accompanying text.
• Minor-Specific Protection: The AI features specialised facial recognition trained to identify underage victims. In cases involving minors, the system is authorised to initiate immediate "preemptive deletion" requests without the standard consent hurdles.
• Worker Welfare: In an effort to protect the mental health of human operators, the system automatically blurs explicit content, ensuring that staff only see what is necessary for final verification.

Scaling the Defense

The effectiveness of the system is reflected in the city's data. Since the AI began its pilot phase, the number of successful deletions handled by the Seoul Digital Sex Crime Support Centre surged from roughly 2,500 in 2022 to over 15,000 in 2025.

"Digital sex crimes are a borderless threat that requires a borderless response," said a spokesperson for the Seoul Metropolitan Government. "By distributing this technology for free, we are removing the financial barriers that prevent smaller municipalities and NGOs from protecting their citizens."

Global Integration

With the official unveiling today, March 4, 2026, the city has opened applications for technology transfer. The software is designed to be compatible with various social media APIs and global server standards, allowing for a coordinated, international crackdown on the distribution networks that host such content.

The initiative marks the first time a municipal government has developed and shared high-level forensic AI for public safety on this scale.