[AI Turns Into Weapon, K-Defense Coordinates]① Four Defense Firms Have Fully Equipped with AI, But Outdated Certification Holds Back K-Defense
Verification Standards Remain Manned-Centric
Hindering Actual Force Implementation
Although domestic defense companies are accelerating the development of next-generation weapon systems utilizing artificial intelligence (AI), outdated certification and testing regimes have become obstacles to their deployment. While technologies such as AI infrastructure, robotics, autonomous flight, and command and control are rapidly advancing, verification standards still remain tailored to manned weapon systems. Experts point out that to turn civilian-developed technologies into actual military assets, new certification systems accommodating AI's autonomous decision-making must be established.
From Infrastructure to Command and Control... Four Defense Companies Arm Every Battlefield Domain with AI
According to the defense industry on July 29, Hanwha, Hyundai Rotem, Korea Aerospace Industries (KAI), and LIG D&A are each enhancing their AI capabilities in core areas such as data infrastructure, unmanned robotics, autonomous flight, and command and control. While each company focuses on different areas, they share a common strategy: to make individual weapon systems intelligent and connect them into a single battlefield system. The speed of AI validation and deployment is considered a key factor that will determine future competitiveness in the defense industry. However, unlike technological development, the institutional groundwork for testing and operational deployment remains insufficient.
Hanwha is focusing on building the infrastructure to support the operation of AI weapon systems. On July 3, Kim Dongkwan, Vice Chairman of Hanwha Group, announced at the National Briefing on the Vision for Advanced Industry Development in Yeongnam that the company plans to invest 55 trillion won in aerospace and AI by 2040. Hanwha Systems is pursuing a space AI data center for real-time storage and analysis of information collected by low-orbit satellites and is also building a defense AI data center in Changwon, Gyeongnam Province. By dispersing data bases in space and on the ground, the company aims to enhance operational sustainability. The goal is to link the acquired data to the multipurpose unmanned vehicle ArionSmet and to the next-generation infantry fighting vehicle (K-NIFV) to support threat analysis and engagement decisions. However, there are still only limited regulations that allow the use of military operational data for civilian technology development and validation.
Hyundai Rotem is concentrating on unmanned robotic platforms that can translate AI decisions into real-world missions. Recently, the Defense Solution Business Division was reorganized into the AD (Aerospace·Defense) & RH (Robotics·Hydrogen) Business Division, and robotics teams from the defense and railway sectors were integrated into the AX Robotics Division. Based on multipurpose unmanned vehicle HR-Sherpa and multilegged walking robots, the company is expanding applications to surveillance and reconnaissance, transportation, and firefighting. In May, Hyundai Rotem was awarded a national project to develop an integrated control system capable of operating multiple robots simultaneously via natural language commands. The company is transitioning from developing individual robots to technologies that can manage swarms of unmanned systems, thereby expanding its business scope. While robot development is accelerating, the standards for testing and evaluating the autonomous decision-making and integrated operation of multiple unmanned systems are still unclear.
While the role of unmanned robots increases on the ground, AI in the air is evolving to substitute part of pilot duties. KAI is developing autonomous flight technology based on the AI pilot “K-AILOT.” The plan is to create the next-generation air combat system (NACS) by connecting K-AILOT with the KF-21, FA-50, unmanned combat aerial vehicles, multipurpose unmanned platforms (AAP), and low-orbit satellites. Having started development in 2023, KAI began actual flight demonstrations with real aircraft this year. The goal is to achieve semi-autonomous formation flight with automatic target identification and avoidance functions by 2028. KAI is also investing in AI specialist companies to broaden the technology ecosystem. However, although the technology is now undergoing flight demonstrations, a certification system to verify the autonomous decision-making of AI has not yet been established.
For each weapon system to make independent decisions and operate on its own, a command system is required to synthesize and control battlefield information. LIG D&A is developing AI-based command and control systems to connect individual forces into a unified network. In July last year, the company began developing next-generation command and control systems applying large language models (LLMs) in partnership with the LG AI Research Institute, aiming to analyze and summarize battlefield data in real time to support commanders’ judgments. LIG D&A also demonstrated technology for commanding multiple unmanned surface vessels from a single system. In collaboration with Palantir, the company is developing an AI decision-support system that reduces the time from detection to engagement to one-tenth of the original. As the scope of AI involvement in a commander’s decision-making expands, so does the need to clarify responsibility and performance verification standards.
Technology Advancing, but AI Deployment Stalled by Outdated Certification Systems
AI technology development and investment in the defense industry have gained full momentum, but the supporting systems and regulations remain stagnant. Weapons systems are rapidly shifting from manned to autonomous and unmanned, yet certification and testing standards are still set for manned aircraft.
KAI exemplifies this issue. While an AI pilot can autonomously enhance its flight and mission implementation skills through reinforcement learning, there is still no airworthiness certification system to verify these capabilities. If current manned aircraft certification procedures are applied, the autonomous decision-making functions learned during demonstration must be re-engineered as rule-based algorithms for actual deployment. As a result, the performance gained through reinforcement learning cannot be utilized, and operations are constrained by human-designed rules. This is why experts warn that accumulated technical achievements may remain stuck at the R&D phase without ever making it into actual deployment.
Such institutional limitations are not confined to KAI. Industry insiders believe that while the K-AILOT has encountered the limits of current systems first, unmanned aerial vehicles, ground vehicles, unmanned surface vessels, and manned-unmanned teaming systems with AI capabilities will face similar challenges. AI-based unmanned systems from Hanwha, robotic platforms from Hyundai Rotem, and command and control systems from LIG Nex1 are all being developed to allow AI to make autonomous decisions and carry out missions. The crucial question of how far to permit AI’s autonomous judgment and by which standards to verify it has now emerged as a shared challenge for the domestic AI defense industry.
Since last year, the industry has been requesting that the government establish a defense AI regulatory sandbox that allows testing of AI algorithms in restricted airspace, as well as a joint military-civilian AI testbed leveraging military operational data. However, related systems have not yet been established. There are concerns that if regulations do not keep pace with technological advancements, the AI competitiveness of domestic defense companies may also weaken.
Hot Picks Today
"Two Spend 500,000 Won in a Flash, Yet Wait Two Hours?" ... 'Dreamland' Shunned by Kids and Foreign Tourists
- The Era of Getting 200 Million Won After 30 Years Is Over... A Japanese Bank's "Perpetual Department Head" Who Just Clocked In Cries Out
- Spain Hit by Its Worst Tragedy... Lee Kangin, Moreno and Real Madrid Join Soccer World in Mourning
- They Said, “Make It Your Home for Life” - Then Suddenly Kicked 200 Passengers Off... What Happened to the Around-the-World Cruise?
- "130 Rice Bowls Sold Out in a Day"... 'Unfiltered Face' Adds 8 Million Won to Night Market Sales: China's Bungeunwoo
Jang Wonjun, Professor at the Department of Advanced Defense Industry at Chonbuk National University, commented, "Domestic defense companies are proactively investing in the development of AI weapon systems, but as long as airworthiness and other certification standards remain unclear, R&D achievements are not translating into actual deployment. The government should promptly establish certification and testing standards appropriate for AI weapon systems so that companies can verify their technologies based on predictable benchmarks."
© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.