The Hong Kong Monetary Authority (HKMA) is exploring the use of artificial intelligence to detect money laundering involving cryptocurrencies. Specialized AI agents could simultaneously monitor blockchain activity and traditional banking transactions, identifying suspicious transfers and helping authorities freeze criminal funds more quickly.
This was announced by Raymond Chan, the HKMA's Executive Director for Enforcement and Anti-Money Laundering. According to Chan, the regulator is exploring the development of systems capable of independently analyzing large volumes of transactions and detecting complex money movement schemes.
In particular, AI could track repeated conversions between cryptocurrencies and fiat currencies, which criminals use to conceal the origins of their funds. The proposed approach would combine blockchain analytics, bank customer information, fraud intelligence and the monitoring of suspicious crypto wallets.
The main challenge remains identifying the owners of cryptocurrency wallets. Although all transactions on public blockchains are permanently recorded and available for analysis, determining who actually controls a particular address is often difficult. Meanwhile, criminals can move stolen assets across multiple platforms and transfer them out of the country within minutes.
To address this problem, the HKMA is already exploring and testing decentralized digital identity technologies. These could help financial institutions identify the parties involved in suspicious transactions more quickly and take action before the funds disappear.
Stablecoins under increased scrutiny
Hong Kong's regulator is paying particular attention to stablecoins. Their relatively stable value and the ability to transfer funds directly between wallets make these assets convenient for both ordinary users and criminals.
According to research by Chainalysis, cryptocurrency addresses linked to illicit activity received at least $154 billion in 2025, an increase of 162% from the previous year. Stablecoins accounted for 84% of total illicit cryptocurrency transaction volume. However, the share of illicit activity in overall cryptocurrency transaction volume remained below 1%, according to the company.
Hong Kong is already strengthening cooperation between banks, cryptocurrency exchanges and law enforcement agencies. A dedicated Virtual Asset Intelligence Taskforce (VAIT) brings together police, customs authorities, financial regulators and licensed cryptocurrency companies. Participants share information about suspicious transactions, identify fraudulent schemes and coordinate asset freezes.
At the same time, the HKMA is developing Project SAGE, launched on November 2. The initiative aims to establish rules for the safe use of autonomous AI agents in the financial system. In particular, experts are examining how to verify the authority granted to such programs, monitor their activities and determine responsibility for potential errors.
In addition, Hong Kong authorities intend to introduce legislation by the end of 2026 that will expand mandatory licensing requirements for cryptocurrency businesses. The new rules will cover companies engaged in digital asset trading, custody, asset management and investment advisory services. Hong Kong already has separate licensing regimes for cryptocurrency exchanges and stablecoin issuers.
Cybercriminals are also turning to AI
Artificial intelligence is increasingly being used not only by banks and regulators but also by criminals themselves. According to an Anthropic study published in June 2026, of 832 accounts blocked for malicious cyber activity, 560 (67%) had used AI to develop malware. In September, the company reported cases in which hackers used AI agents to identify vulnerabilities, infiltrate corporate systems and steal data. Some attacks were carried out almost automatically, with minimal human involvement.
In the cryptocurrency industry, AI is also helping fraudsters deceive users and steal funds. According to Chainalysis, AI-enabled scams generated 4.5 times more revenue for criminals than traditional scams in 2025. Criminals use deepfakes, fake identities and automated tools to identify potential victims. These technologies allow fraudulent operations to be conducted on a much larger scale and make criminal schemes more difficult to detect.
As a reminder, Hong Kong is launching round-the-clock settlement for tokenized assets.