Korea bolsters AI capabilities to tackle crypto market manipulation

gettyimagesbank

The Financial Supervisory Service (FSS) is upgrading an artificial intelligence (AI)-powered platform for virtual asset surveillance as part of efforts to crack down on increasingly sophisticated market manipulation, the watchdog said Monday.

The platform, Virtual Assets Intelligence System for Trading Analysis (VISTA), is designed to process massive volumes of trading data, automatically flag abnormal transactions and visualize trading behavior. The financial watchdog said VISTA has already become a core investigative tool in crypto-related enforcement cases.

Last December, FSS expanded its computing capacity by installing two new servers equipped with high-performance central and graphics processing units.

Leveraging this infrastructure, the agency has developed an automated detection model that identifies suspected price-rigging periods without manual intervention.

The system scans every possible trading interval — from seconds to months — and calculates abnormal trading metrics across all segments, allowing authorities to uncover manipulation patterns regardless of how frequently they occur.

“We will expand AI-based analytical capabilities in stages through the end of this year to detect illegal trading earlier and accelerate investigations,” an FSS official said. “To that end, a 170 million won ($116,000) budget has been secured for 2026 to upgrade servers and enhance the performance of AI surveillance systems.”

The plans include rolling out new tools designed to automatically identify networks of suspicious accounts involved in coordinated price manipulation. At the same time, the watchdog aims to introduce analytics capable of processing and interpreting abnormal trading-related text across thousands of crypto assets.

In a separate move, the Financial Services Commission (FSC) and the Korea Exchange said they will launch an AI-driven cyber surveillance system on Tuesday to strengthen early detection of stock price manipulation and other market abuses.

The system was developed in response to a rise in online misconduct, including the spread of false information and coordinated stock promotion through online forums, social media platforms and YouTube in attempts to exert undue influence on stock prices.

It analyzes digital content such as posts, spam text message reports and videos alongside stock price movement data. Authorities will review securities flagged by the system and conduct in-depth investigations when necessary.

“We will continue expanding the use of AI to combat unfair trading in capital markets,” an FSC official said.