Naver Cloud Proposes Defense AI Strategy for Rapid Military Technology Integration

By Shin Hye An Posted : September 18, 2026, 11:12 Updated : September 18, 2026, 11:12


Naver Cloud hosted the 'Defense AI Strategy Seminar' on September 18, presenting a strategy for the military to rapidly integrate civilian technology while maintaining control over data and artificial intelligence (AI) models.

The event, held at the Shilla Hotel in Seoul, attracted 550 participants, including officials from the Ministry of National Defense, the Joint Chiefs of Staff, and various military branches. Naver Cloud emphasized the need to accelerate the transition to defense AI by consolidating, connecting, validating, and evolving civilian technologies.

Kim Yoo-won, CEO of Naver Cloud, stated, "Defense AI should not end with technology adoption; it must define the problems the military needs to address and validate them in the field for effective integration." He added, "Our role is to support the military with the fastest and most accurate technology in the areas they require."

He further noted that Naver Cloud aims to be a key partner in connecting and supporting the capabilities of the military, defense companies, startups, and research institutions to solve real defense challenges.

Naver Cloud outlined its strategy to go beyond being a mere AI provider, positioning itself as a long-term partner in designing and validating the AI transition for South Korea's defense. The company plans to take responsibility for everything from infrastructure to actual mission applications, leveraging its full-stack capabilities and proven security technologies.

Former Deputy Minister of National Defense Kim Sun-ho emphasized that the success of Defense AX hinges on how quickly the military can convert civilian innovative technologies into operational capabilities. He stressed the necessity of establishing a 'civil-military fusion ecosystem' under the principle of 'civilian technology first,' where the private sector leads technological innovation and the military swiftly integrates it.

He also highlighted the importance of not just validating technologies but also organically integrating AI into military operational concepts and weapon systems to achieve practical deployment and dissemination.

Yoo Kyung-beom, Executive Vice President of Naver Cloud's Defense Business, proposed that the acceleration of sovereign AI in defense requires a structure that consolidates, connects, validates, and evolves civilian technologies.

Naver Cloud plans to gather expertise from various civilian sectors, including infrastructure, AI models, and applications, ensuring that military data is securely connected to AI models in a closed network environment, allowing the military to maintain control over data and models. The company intends to validate results progressively in mission areas where effectiveness can be measured and continuously evolve performance in collaboration with Field Deployment Engineers (FDE).

Sung Nak-ho, who leads model development at Naver Cloud, suggested that for autonomous weapon systems to function correctly, a 'common semantic framework' is necessary to ensure that sensors and platforms can understand the same situations.

He proposed technologies such as spatial intelligence, which structures drone footage and sensor information in three-dimensional space, and world models that simulate and predict outcomes in virtual environments reflecting battlefield variables like terrain, weather, and radar.

In the future, Naver Cloud aims to connect the entire process of situational understanding, judgment, action, and learning, allowing the military to directly control and implement the intelligence they desire without being dependent on specific models.

Naver Cloud's strategy is to provide ongoing support rather than one-time technology supply, continuously enhancing performance by reflecting real-time field requirements through FDE. This approach positions the company as a key technology partner in enabling the military to directly control data and models, ultimately creating a self-evolving operational system.





* This article has been translated by AI.

Copyright ⓒ Aju Press All rights reserved.