HK rolls out Fintech 2030 to push AI, tokenisation

The Hong Kong Monetary Authority (HKMA) plans to work with the Insurance Authority and Securities and Futures Commission to promote fintech, including AI adoption, over the next five years under its fintech 2030 strategy.

Fintech 2030, which began this year, will expand the city’s generative AI sandbox to more industries, aiming to improve risk management, fraud detection, and customer experience.

The HKMA is also developing Project Noor, an AI model auditing tool for banks, in collaboration with the Bank for International Settlements Innovation Hub and the UK Financial Conduct Authority.

Tokenisation is another focus, with banks and fund houses expected to use tokenised deposits and explore real-time settlement for financial transactions.

🔗 Source: South China Morning Post

🧠 Food for thought

Implications, context, and why it matters.

Hong Kong Monetary Authority (HKMA) timeline gaps leave banks unsure about compliance or trials

HKMA’s five-year Fintech 2030 plan lists over 40 moves 1. Public materials set no dates for GenAI Sandbox expansion or use of the Commercial Data Interchange (CDI), and do not say what is optional versus binding.
The GenAI Sandbox launched in September 2024 2 for Authorized Institutions (AIs), which are Hong Kong-licensed banks and deposit-taking companies. Nonbanking access has no stated eligibility, application steps, or go-live timing. Project Noor offers supervisory tools for transparency, fairness, and robustness with privacy 3, and public material casts it as a support kit rather than a mandatory audit system 4, which leaves banks to pick between building explainability now or waiting. That creates risk of overspending or late compliance.

AI governance vendors can align with Project Noor supervisory toolkit

Noor gives regulators explainable AI tools to check transparency, fairness, and robustness while preserving privacy 3. The push aligns with rules such as the European Union AI Act on explainability, and a 2024 survey in external analysis says only 34% of financial firms fully understand how their AI works 5, so compliance-as-a-service firms (subscription tooling and expertise) can fill the gap.
Audit and governance platforms should align roadmaps to Noor pillars and target banks in Hong Kong, the United Kingdom, and European Union 3. Vendors need details on architecture and pilots to time launches, yet public docs stay high level.