Hong Kong Monetary Authority: Concerned about how banks can effectively manage AI will issue guidelines on practical application

The Zhitong Finance App learned that Hong Kong Monetary Authority Vice President Yuen Kwok-hang said in his speech at the Hong Kong Bankers Summit that artificial intelligence is no longer just an aid in decisions and actions, but can be implemented. He is concerned about how banks can effectively govern AI, improve technological resilience and ensure that banking systems evolve in sync with an increasingly complex environment, how to build a trustworthy tokenization framework, and ensure data quality. At the same time, he said that the Hong Kong Monetary Authority will issue relevant practical application guidelines to help banks translate the principles of data excellence into concrete actions.

Ruan Guoheng pointed out that AI can provide decision-making solutions and execute some processes, but this does not mean that every application scenario must be automated or autonomous. He emphasized the importance of human supervision, pointing out that in the end, it is still up to people to make decisions and take responsibility. Tokenization makes the transfer of value programmable, changes the transaction construction process and how financial services are delivered to users, and creates many opportunities for banks. However, he stressed that trust must not be sacrificed; it is necessary to ensure that all transactions made through blockchain are true, accurate, and authorized.

He stressed that the Hong Kong Monetary Authority still believes that the cutting-edge AI model has not fundamentally changed the face of cyber risk management, but that standards should be adjusted to deal with more rapid and large-scale cyber attacks. The authorities actively guide the industry to deal with related risks. They also expect banks to strengthen cybersecurity controls and assess whether services provided by third parties have taken sufficient risk management measures, and whether they will create a risk of backdoor attacks.

He mentioned that the Hong Kong Monetary Authority launched a generative AI sandbox a few years ago to test more than 40 use cases, half of which have already been commercialized.