Korea’s AI Ecosystem Advances and Secures Agentic AI With NVIDIA
Korean AI firms are integrating NVIDIA’s OpenShell runtime and Nemotron datasets to enhance secure agentic AI deployment across enterprises and robotics.
Hyundai Motor Group is testing NVIDIA’s NemoClaw blueprints and OpenShell runtime to support secure AI agents, while AIM Intelligence is linking its Starfort platform with OpenShell for employee productivity and enterprise workflows. OpenShell, part of the NVIDIA Open Agent Safety Platform, provides sandboxed execution and policy-based controls to protect sensitive data and govern agent actions. The collaboration aims to combine OpenShell’s access controls with Starfort’s detection of malicious instructions and data leakage.
AIM Intelligence is also exploring a preliminary proof of concept in the public energy sector, with potential applications extending to enterprise, industrial, and robotics environments. The integration seeks to enhance security for AI agents handling critical workflows by combining runtime safeguards with threat detection capabilities. OpenShell’s policy-based controls allow enterprises to define granular permissions for agent interactions with tools and data.
LG AI Research is opening API access for K-EXAONE-2.5, a model developed using NVIDIA Nemotron datasets, including Nemotron-Personas-Korea. Nemotron-Personas-Korea is an open dataset of 7 million fully synthetic Korean personas, grounded in official Korean statistics, to ground training data and evaluation in Korean contexts. The model and dataset are designed to support developers in building specialized, efficient AI agents tailored to Korean language and cultural contexts.
NVIDIA Nemotron is a family of open models, datasets, and techniques that help developers build specialized, efficient AI agents. The initiative reflects a broader effort to advance agentic AI in Korea by combining secure runtime environments with culturally grounded datasets and open model access.