HKMA Launches Second GenA.I. Sandbox Cohort to Enhance A.I. Security

The Hong Kong Monetary Authority (HKMA), in partnership with the Hong Kong Cyberport Management Company Limited, has unveiled the second cohort of its Generative Artificial Intelligence (GenA.I.) Sandbox. This initiative is aimed at advancing secure and responsible A.I. innovation within the financial industry, according to the Hong Kong Monetary Authority.

Expanded Participation and Innovative Use Cases

The latest cohort features 27 innovative use cases selected from over 60 proposals submitted by 20 banks and 14 technology partners. The submissions were evaluated based on innovation, technical complexity, and potential industry value. This marks a significant shift from A.I. experimentation to its practical, secure implementation.

Focus on A.I. Governance and Fraud Prevention

A key focus of this cohort is proactive A.I. governance, exemplified by the pioneering use of “A.I. vs. A.I.” strategies. These strategies involve using A.I. to perform automated quality checks on A.I.-generated outputs, thereby enhancing accuracy and consistency. In addition, the Sandbox will address deepfake-related fraud by serving as a testing ground for advanced defense mechanisms against digital fraud.

Industry Collaboration and Future Developments

Mr. Arthur Yuen, Deputy Chief Executive of the HKMA, highlighted the cohort‘s role in making A.I. adoption safer and more robust, underscoring the industry-wide consensus on A.I.’s transformative potential. The initiative reflects growing collaboration between the banking, technology, and academic sectors, contributing to the development of Hong Kongs fintech ecosystem.

Participants will begin onboarding to Cyberports Artificial Intelligence Supercomputing Centre later this year, with trials set to commence in early 2026. The HKMA plans to leverage the GenA.I. Sandbox to promote responsible A.I. adoption and share best practices within the financial sector.