The Hong Kong Monetary Authority (HKMA) is taking steps to support the use of artificial intelligence (AI) in enhancing the monitoring of money laundering and terrorist financing (ML/TF) risks within authorised institutions.
A dedicated team, backed by an external consultant, will offer supervisory feedback and technical guidance to help institutions implement AI-driven monitoring processes through the HKMA’s existing Fintech Supervisory Sandbox and Chatroom.
The HKMA had previously shared a list of high-level principles on AI with authorised institutions in 2019.
Given the rising complexity of money laundering activities and the growing use of mule account networks, the HKMA encourages institutions to strengthen their monitoring systems.
AI-powered solutions have already shown better efficiency and effectiveness compared to traditional rules-based methods by utilising a broader range of contextual information to flag suspicious activities.
These systems also leverage past transaction patterns and customer risk profiles to enhance their effectiveness.
The HKMA plans to further foster the adoption of AI by sharing success stories from the industry at a forum in November 2024.
The regulator will also continue to provide targeted support for institutions as they explore AI’s potential for improving anti-money laundering and counter-financing of terrorism (AML/CFT) efforts.
These authorised institutions are advised to conduct feasibility studies on integrating AI into their systems and submit an implementation plan to the HKMA by March 2025.
This study and the implementation plan must be signed off at the board level before submission.
Raymond Chan, Executive Director (Enforcement and AML), said in a circular,
“The HKMA will communicate with AIs [authorised institutions] on an individual basis regarding the exact timing for the feasibility study and implementation plan and the format in which they should be provided, and will consider further engagement and follow up in due course.”
Featured image credit: Edited from Freepik
