South Korean Tech Giants Partner With NVIDIA to Secure and Scale Agentic AI
Hyundai Motor Group, AIM Intelligence, and LG AI Research are deploying NVIDIA’s OpenShell runtime, NemoClaw blueprints, and Nemotron datasets to secure autonomous enterprise agents and advance localized Korean models.
South Korean Tech Ecosystem Adopts NVIDIA Agent Frameworks
As autonomous AI agents take on increasingly complex operational responsibilities, tech leaders in South Korea are prioritizing secure execution environments and context-aware foundation models. Enterprise giants and security specialists in the country—including Hyundai Motor Group, AIM Intelligence, and LG AI Research—are leveraging NVIDIA’s open agent frameworks, safety platforms, and specialized datasets to advance the nation's agentic AI deployment.
The initiatives target the end-to-end lifecycle of enterprise agents, addressing sandboxed execution, protection against malicious instructions, and localized model training.
Hyundai and AIM Intelligence Fortify Enterprise Agents With OpenShell
Hyundai Motor Group is exploring NVIDIA NemoClaw blueprints alongside the NVIDIA OpenShell runtime to support secure enterprise agents tailored for workplace productivity and operational workflows.
Serving as a core component of the NVIDIA Open Agent Safety Platform, OpenShell offers a secure runtime environment using sandboxed execution and policy-based controls. The platform is designed to govern autonomous agent actions, control access to corporate tools, and safeguard sensitive internal data during real-time tasks.
In parallel, South Korean AI security company AIM Intelligence is integrating its Starfort agent control platform with NVIDIA OpenShell. The joint system combines OpenShell’s runtime access controls with Starfort’s threat detection capabilities, which identify malicious instructions and prevent data leakage. AIM Intelligence is evaluating a preliminary proof of concept for the public energy sector, with potential future deployment across industrial, enterprise, and robotics environments.
LG AI Research Opens API Access for K-EXAONE-2.5
Complementing agent runtime security, LG AI Research is expanding public model access by opening application programming interface (API) availability for its K-EXAONE-2.5 model.
K-EXAONE-2.5 was developed using NVIDIA Nemotron open datasets, specifically incorporating Nemotron-Personas-Korea. This open dataset contains 7 million fully synthetic Korean personas created from official South Korean statistics, enabling developers to ground model training and evaluation directly within Korean socio-cultural contexts.
The datasets belong to the broader NVIDIA Nemotron family—an open collection of models, synthetic datasets, and techniques designed to help developers build specialized, efficient AI agents.
