Naver Cloud and LG CNS are developing a cybersecurity-focused AI foundation model for South Korea.
The AMW Read
The consortium meaningfully extends South Korea's foundation-model roster into cybersecurity-specific, air-gapped deployments with a broad domestic institutional coalition.
Naver Cloud and LG CNS are developing a cybersecurity-focused AI foundation model for South Korea.
Naver Cloud and LG CNS have formed a consortium for a Ministry of Science and ICT and National IT Industry Promotion Agency program to develop a model focused on detecting and responding to domestic cyber threats. The project aims to run in air-gapped environments using domestic technology across the model, infrastructure, security, and data layers. Naver Cloud will provide the model and cloud infrastructure, drawing on deployments in nuclear and financial closed-network environments, while LG CNS contributes red-team, blue-team, and AI-security specialists. The 33-member consortium also includes security companies, universities, research bodies, and public institutions.
This is a practical sovereign-model effort centered on a high-stakes deployment constraint: security data must remain inside controlled environments. The consortium is not simply adapting a general chatbot for security operations; it is targeting locally observed attack patterns and domestic security environments, while incorporating threat-monitoring and defense experience from Naver services. That approach makes domain data, operational testing, and controlled deployment as important as the base model itself. Per the AI Market Watch index, Naver Cloud name-matched coverage rose to four items in the last 90 days from two in the prior 90, limited to pipeline-ingested sources.
For builders, the implication is that enterprise security AI will need credible deployment architectures alongside detection capability, especially for regulated and isolated networks. For investors, the more durable value may accrue to vendors that pair specialized models with proprietary security telemetry, validated red- and blue-team workflows, and the ability to deploy on premises without exposing sensitive data to external services.