Korean Companies Partner with NVIDIA to Advance Agentic AI with Security

Image: NVIDIA · Source

Hyundai and AIM Intelligence in Korea are integrating NVIDIA’s Nemotron models and OpenShell runtime to enhance secure, capable AI agents for enterprise and industrial applications.

As artificial intelligence agents play increasingly complex roles, Korean technology companies are collaborating with NVIDIA to develop more capable and secure AI systems. Hyundai Motor Group is exploring NVIDIA’s Nemotron open models and the OpenShell runtime environment. Simultaneously, AIM Intelligence is integrating its Starfort platform with NVIDIA OpenShell to create secure AI agents designed to enhance employee productivity and streamline enterprise workflows.

NVIDIA OpenShell is a core component of the Open Agent Safety Platform. It provides a secure runtime featuring sandboxed execution and policy-based controls. These measures protect sensitive data, regulate access to enterprise resources, and govern autonomous agent behaviors to ensure safe AI operations.

AIM Intelligence, a Korean AI security firm, is combining the runtime access controls of OpenShell with its Starfort platform’s capabilities to detect malicious instructions and prevent data leakage. They are also exploring a proof of concept within the public energy sector, with intentions to extend these secure applications across enterprise, industrial, and robotics settings.

On the AI model front, LG AI Research is opening API access to K-EXAONE-2.5, an AI model developed using NVIDIA’s Nemotron datasets — specifically Nemotron-Personas-Korea. This dataset includes 7 million fully synthetic Korean personas, grounded in official statistics, enabling model developers to firmly anchor their training and evaluations in relevant Korean contexts.

NVIDIA Nemotron encompasses open models, datasets, and methodologies that assist developers in building specialized, efficient AI agents. These initiatives illustrate a significant step towards advancing the Korean AI ecosystem, focusing on both capabilities and security as AI agents undertake more autonomous tasks.

For developers and enterprises interested in building secure AI agents, NVIDIA provides resources and tools through the Nemotron models and the OpenShell runtime platform.