The HKMA did not specify the eligible requirements for the GenAI sandbox and its application process, but said these details would be announced in due course.
The latest collaboration between the HKMA and Cyberport reflects how GenAI is becoming a critical tool for financial institutions seeking to enhance their operations and customer services.
“This new sandbox aims to overcome the ‘hard’ and ‘soft’ barriers to the adoption of GenAI – that is, the demand for computing capabilities and the need for supervisory guidance,” HKMA executive director Carmen Chu Lap-kiu told the audience at the launch during the FiNETech2 event held at Cyberport.
“As AI systems become more sophisticated and influential, we must ensure their ethical development and deployment, and the core principle of responsible innovation should also come to the forefront of considerations.”
While GenAI is expected to have a significant impact across various sectors, banking, hi-tech and life sciences are among the industries that could see the biggest impact, according to a June 2023 report by consultancy McKinsey on the technology’s economic potential.
“Across the banking industry, for example, the technology could deliver value equal to an additional US$200 billion to US$340 billion annually if [GenAI’s] use cases were fully implemented,” the McKinsey report said.
Hong Kong’s financial institutions are ahead of other markets in embracing generative AI in their daily work, according to an annual financial services survey, conducted by financial software provider Finastra in November 2023. This survey found about 38 per cent of Hong Kong finance executives have started rolling out GenAI, the highest of all the markets surveyed and above the global average of 26 per cent.
GenAI can be used to automate existing processes and controls, which is expected to improve efficiency and also cut costs, according to consultancy KPMG’s Hong Kong Banking Report 2024. This report also said the city’s banks must remain cautious about using GenAI on customers’ data in light of privacy and governance concerns.