🔗 LG CNS, Security AI Strategy Faces the Test of Business Expansion - 2026.09.10

Hi, it’s Dotori. The key point this time is that LG CNS is trying to elevate security AI from a simple technology development task to a pillar of Korean-style digital defense capabilities. The company has laid out a direction to develop Korean-style security AI while pointing out the limits of relying on overseas security technologies, and the stage for its existing generative AI business is expanding as the effort connects to government projects and digital defense. ^[1]

🔗 Let’s start with the big picture

LG CNS believes there are limits to relying on overseas security technologies and has presented a direction to develop “Mythos-level” Korean-style security AI. ^[1] What matters here is that this is not simply about creating one more new AI model; it reveals a business direction aimed at securing technology suited to the domestic environment in the sensitive area of security.

Looking one step further, the stage the company is discussing does not stop at enterprise security. It has also mentioned the challenge of expanding security AI business to strengthen digital defense capabilities. ^[2] The requirements for business automation at private companies and security demand in defense and public sectors inevitably differ, so the data and validation methods needed during deployment may also vary even for the same AI. That connection is the key. It means another step is being placed between model development and actual field deployment.

⚙️ What the performance target shows

In relation to the government project to develop a security-specialized AI foundation model, Eun-sook Jeon, head of LG CNS’s security business, set a target of achieving 100% of the performance of global models. ^[3] However, this figure must be viewed separately as a target, not an achieved result. It is an indicator showing the technology direction, not proof that performance has already been verified.

So the question becomes simple. It is not enough to aim for the same level as global models; can this lead to a model that is actually usable in Korea’s security environment? Unlike a general document-writing tool, security AI is difficult to evaluate through a demo screen alone, and field deployment and validation are likely to have a major influence on the weight of the business. Of course, this is an interpretation based on the goals reported in the article and does not mean that actual results or monetization have been confirmed.

🏭 Which part of the value chain will move first?

If we separate the broader industry picture from the value chain, the first priority in this news can be seen as security AI model development, the second as deployment in defense and public sectors, and the third as expansion across the generative AI and security AI businesses overall. The structure is that the next cog, defense and public-sector operations, can mesh only after the security AI model cog starts turning.

However, the stages may not move at the same speed. Even if model development progresses, the requirements of the adopting institutions, validation procedures, and operating environments must align for it to lead to an actual business. Conversely, as field deployment cases accumulate, it may become clearer that LG CNS’s security AI is moving beyond a simple development task and taking root as a business area. Dotori believes the central point for the next judgment is whether security AI leads to actual field data and operational results.

📚 The remaining variable is ‘what comes after the target’

What has been confirmed this time is the direction to develop Korean-style security AI, the willingness to expand the business aimed at strengthening digital defense capabilities, and the target of 100% of global model performance. ^{[1] [2] [3]} On the other hand, the articles alone do not allow us to confirm whether the target has been achieved, the specific business scale, or its contribution to revenue.

Ultimately, LG CNS’s security AI story sits at the point of transition from a technology blueprint to a field business. The remaining questions are how model performance will be validated within government projects, whether actual deployment cases will follow in defense and public sectors, and how security AI will combine with the company’s AX business. Whether the security AI cog will connect all the way to the gears of real industrial sites is likely to become clearer in the next scene.

📰 Reference News (3)

1

[Interview] “There are limits to relying on overseas technology”… LG CNS to develop ‘Mythos-level’ Korean-style security AI - ZDNet Korea

지디넷코리아 · 2026.09.10

📰 View original

2

[Meeting] “We must strengthen digital defense capabilities”… LG CNS challenges for No. 1 in ‘security AI’ - yna.co.kr

yna.co.kr · 2026.09.10

📰 View original