The number of money laundering cases involving cryptocurrencies in South Korea increased 152-fold in the first half of 2026: from eight cases for all of 2025 to 1,214 from January to June of the current year. This data was provided by the country's National Police Agency. The share of money laundering in the total number of detected cryptocurrency-related crimes reached 79.4% — compared to 92% a year earlier for investment fraud, which was previously the main category.
The "Hwanchigi" Scheme
The primary tool remains the Hwanchigi technique — it allows the withdrawal of illegal funds from the country through cryptocurrency transfers, bypassing South Korea's regulated banking system. This method is fast, cross-border, and almost impervious to prosecution. A Crystal Intelligence report linked the overwhelming majority of the country's illegal crypto flows from 2021 to August 2025 to this scheme, and fresh police statistics point to a sharp acceleration in its use.
South Korean authorities have been tightening control over illegal crypto operations for several years — both law enforcement and regulators have strengthened measures. However, case data shows that criminal networks are scaling up faster than the law enforcement system can keep up.
Tether as the Primary Tool
The preferred means for laundering remains USDT. Stablecoins now dominate the structure of illegal cryptocurrency transactions worldwide, and South Korean criminal groups use them to convert proceeds from drug trafficking, illegal gambling, and phishing into dollars, then withdraw the funds through overseas exchanges outside the country's jurisdiction.
Detection Exists, Punishment Does Not
The gap between detection statistics and real consequences for criminals looks telling. In the first half of 2026, police made only 18 arrests in cryptocurrency-based money laundering cases — compared to 42 for all of 2023, while the number of detected cases increased nearly 100-fold.
This pattern repeats in recent high-profile operations. In June 2026, the Seoul Metropolitan Police indicted 23 members of a network linked to a Cambodia-based phishing gang and seized $431,000 in illegal proceeds. However, the alleged mastermind of the scheme remains at large with an Interpol "red" notice.
In July 2026, investigators tracked and blocked $12 million in XRP and Tether after a fake staking website impersonating Flare Network siphoned $8.6 million from 71 investors. Arrests in this case significantly lagged behind the asset freeze.
Billions Through Illegal Channels
South Korea's Customs Service seized 7.2 trillion won (approximately $4.92 billion) in the first half of 2026 in cases of illegal foreign exchange operations — including those involving export companies that accepted cryptocurrency to bypass revenue repatriation rules. Over 90% of the 9.5 trillion won in cryptocurrency-related crimes referred for prosecution passed through unlicensed channels, not regulated banks.
South Korean authorities are capable of tracking fund movements almost in real-time. Bringing a case to court is a completely different challenge.
AI Opinion
From a machine data analysis perspective, the Hwanchigi scheme does not appear to be a unique South Korean invention — structurally, it repeats the practice of "mirror transactions" that Chinese money laundering networks have been using for years to move funds across borders. Such methods cost only 2-5% in fees compared to 17% for professional launderers — the low entry price explains the speed of scaling such schemes in any jurisdiction.
The situation demonstrates a systemic gap between the speed of blockchain analytics and the slowness of the judicial machinery: transactions are tracked almost instantly, while charges are formed over months. A similar asymmetry has already manifested in cases involving the Blender and Sinbad mixers, where sanctions arrived years after the first signals. An open question remains: can detection statistics even serve as an indicator of anti-money laundering effectiveness if it does not change the economics of the crime itself?
end-content
