NTT Com Asia has released a service blueprint for its Financial AI Fabric in Hong Kong.

The Financial AI Fabric blueprint enables financial services institutions (FSIs) to move AI workloads from sandbox environments into full-scale production through a private AI infrastructure layer designed for data sovereignty and regulatory compliance.

The company debuted the blueprint at the LEAP East 2026 tech conference.

According to NTT Com Asia, financial institutions are increasingly moving away from cloud-only setups toward hybrid models. These models combine public cloud, colocation facilities and sovereign data centres to support AI workloads.

Core infrastructure components

The Financial AI Fabric addresses this demand through three interconnected components. The first is a photonic network, APN InterLink, which provides ultra-low latency data exchange for distributed AI workloads.

The second involves AI data centres equipped with direct liquid cooling technology and an AI-driven smart building platform to handle high-density power constraints.

The third is a sovereign GPU compute environment designed for secure private AI adoption.

“By establishing this service roadmap, we are reinforcing Hong Kong’s position as a leading financial AI hub and providing the entire FSI ecosystem with a clear, compliant path toward secure, sovereign private AI,”

Daisuke Kuroda, Chief Executive Officer of NTT Com Asia, said.

Hong Kong digital infrastructure

To support the blueprint, NTT Com Asia operates two financial-grade AI data centres in Hong Kong: the Tseung Kwan O Financial Data Center and the Tai Po Data Center.

The company commercially launched its APN InterLink infrastructure in the city in November 2025.

NTT Com Asia also plans to introduce a GPU-as-a-Service offering later this year. The service will provide regional institutions with localised compute resources on demand.

Featured image credit: Edited by Fintech News Hong Kong, based on image by