Mizuho Partners with NVIDIA to Build Secure, High-Performance AI Infrastructure for Banking

Mizuho Financial Group has launched an initiative to upgrade its artificial intelligence capabilities, partnering with NVIDIA to explore and develop high-security AI infrastructure tailored for the financial services sector. The collaboration focuses on deploying advanced hardware and secure execution frameworks to integrate generative AI safely into the bank's daily operations.

Dual-Pronged Strategy: Computing Power and Secure Execution

The initiative, led by Mizuho President & Group CEO Masahiro Kihara, focuses on two primary technical tracks designed to balance raw computing performance with the stringent governance required of a global financial institution:

On-Premises Hardware Scaling with DGX B200

Mizuho will integrate NVIDIA’s high-performance DGX B200 platform to strengthen its on-premises GPU capabilities. This environment will serve as the foundation for training, evaluating, and running inference on the bank’s proprietary large language model, "Mizuho LLM," which is being customized with specialized financial knowledge and internal compliance rules. Mizuho is also assessing the architecture needed to scale this system into a multi-server GPU cluster to handle future demand.

Securing AI Agents via NVIDIA NemoClaw

To address the strict regulatory, legal, and auditing benchmarks of the banking industry, Mizuho is conducting technical validation of secure environments using NVIDIA NemoClaw. The trial will test critical security guardrails, including execution environment isolation, network access control, data protection, and the verification of execution histories.

The Bottom Line for Operations

The ultimate goal of the upgrade is to safely transition autonomous AI agents into active business workflows. Once verified, these agents will assist employees with highly confidential tasks—such as information gathering, document preparation, data analysis, and software development support—without risking external data leaks.

By offloading repetitive operational tasks to secure AI partners, Mizuho aims to optimize internal efficiency, allowing its workforce to focus on higher-value client advisory roles and strategic growth.

Mizuho’s Custom AI Matches GPT-5.2 Accuracy on Wall Street Speed, Keeps Bank Data Strictly On-Premise

Mizuho Financial Group has successfully developed a proprietary, finance-specific Large Language Model (LLM) capable of delivering highly accurate, sub-second responses without relying on prolonged AI reasoning processes. Crucially, the new model operates entirely within Mizuho’s secure, on-premise network, allowing the bank to process highly confidential data

Japan FinTech ObserverNorbert Gehrke