Yuen expands AI sandbox to 30 institutions and sets quantum target
Hong Kong Monetary Authority Deputy Chief Executive Arthur Yuen outlined supervisory priorities at the HKIB Annual Banking Conference, unveiling the GenA.I. Sandbox++ with 30 financial institutions and setting a sector-wide quantum readiness target for 2030.
From experimental models to agentic action
The HKMA expanded its generative artificial intelligence initiatives by launching GenA.I. Sandbox++, scaling participation to 36 use cases across 30 financial institutions and 27 technology partners.
Yuen stressed that boards and senior management remain fully accountable for automated agent workflows across the entire decision path.
On infrastructure resilience, the banking sector scored 2.3 out of 10 on its first Quantum Preparedness Index, with an average estimated implementation timeline of 5.6 years.
The supervisor established a full sectoral readiness objective for 2030 and is developing a Post-Quantum Cryptography Transition Toolkit alongside a supervisory incubator for distributed ledger technology.
Point data replaces reporting templates
A supervisory stocktake shows 95 percent fintech adoption among surveyed banks, though 41 percent identify high-quality data as a persistent operational hurdle.
To modernize compliance, the HKMA is introducing Granular Data Reporting 3.0, replacing aggregated templates with point-centric submissions under the principle of reporting once for multiple uses.
This shift complements the transition from the initial 2024 sandbox cohort of 15 use cases into cross-sector agentic workflows.
High ambitions meet fragile foundations
The push exposes severe readiness gaps, highlighted by the dismal 2.3 quantum score and chronic data hurdles.
Deploying autonomous AI agents escalates governance risks beyond what sandboxes can resolve.
Hong Kong lenders must accelerate core IT upgrades to meet the 2030 deadline.