NC AI and Posco DX sign agreement to develop general-purpose robot AI model

NC AI and Posco DX signed a deal to jointly build a general-purpose robot foundation model. NC AI leads the AI development; Posco DX provides testing infrastructure and automation expertise.

Categorized in: AI News IT and Development
Published on: May 31, 2026
NC AI and Posco DX sign agreement to develop general-purpose robot AI model

NC AI and Posco DX Partner on Robot Foundation Models

NC AI, the artificial intelligence subsidiary of NCSoft, signed a strategic agreement with Posco DX on Friday to jointly develop a general-purpose robot foundation model. The partnership targets the creation of AI systems that can serve as the "brain" for various types of robots across industrial applications.

The collaboration focuses on several technical areas: optimizing vision-language-action models, building digital twin simulation environments, and validating robot intelligence systems. NC AI will lead the development and optimization of the VLA model, while Posco DX will provide testing infrastructure and automation expertise.

How the Partnership Works

Posco DX, the IT division of Posco Group, brings decades of robotics and automation experience. The company will establish test environments and help stabilize the technology for operational deployment.

NC AI will focus on the VLA model-a key component for next-generation robot foundation models-and create the digital twin environment where AI models can be validated before real-world use. Testing in a virtual environment allows developers to refine a robot's situational awareness and autonomous decision-making without physical prototypes.

Why This Matters for Development Teams

Foundation models for robotics remain an emerging field. Most robot systems today are built for specific tasks with limited ability to transfer skills across different hardware or environments. A general-purpose robot foundation model could change that.

The digital twin approach is particularly relevant for teams working on AI Agents & Automation. Testing and validation in simulation reduces development cycles and catches failure modes before deployment.

The work also connects to broader developments in Generative AI and LLM architectures. Vision-language-action models extend the same transformer-based approaches used in large language models to robotics, enabling robots to understand both visual input and natural language instructions.

Kim Min-jae, chief technology officer at NC AI, said the companies aim to "lead the general-purpose physical AI ecosystem." Yoon Suk-june, head of the robot automation center at Posco DX, described the partnership as strengthening "automation technologies and internalizing key solutions in robot control and operation."


Get Daily AI News

Your membership also unlocks:

700+ AI Courses
700+ Certifications
Personalized AI Learning Plan
6500+ AI Tools (no Ads)
Daily AI News by job industry (no Ads)