South Korea retools national security policy to counter AI-driven cyber threats

National cyber resilience task force

The South Korean government is taking significant steps to fortify its digital defenses against evolving cyber threats. The National AI Strategy Committee recently announced the creation of a "National Cyber Resilience Strengthening Task Force" to protect the country's critical infrastructure. This initiative aims to ensure that essential services can be maintained and restored quickly if they are targeted by hackers.

This move follows a series of high-profile incidents that have highlighted vulnerabilities in public and private systems. For example, the Korea’s Financial Security Institute recently suffered a hacking attack that resulted in the leak of personal information. Additionally, a data breach at the digital certificate issuer DigiCert exposed enterprise customer contact information after an internal system was left open to the internet.

The government is also looking at specific hardware vulnerabilities, such as those found in internet of things (IoT) products like IP cameras. The Ministry of Science and ICT plans to increase security oversight for manufacturers and distributors of these devices. By creating a centralized task force, the government hopes to provide a unified response to the increasing speed of AI-assisted cyberattacks.

Protecting the cold-chain and finance

Recent international incidents have prompted South Korea to scrutinize its own industrial security. Following reports that Iran was behind cyberattacks on supermarket refrigeration and freezing equipment in Europe, the Ministry of Science and ICT began inspecting the domestic cold-chain industry. The ministry has recommended emergency security measures for companies involved in transporting and storing temperature-sensitive goods.

In the financial sector, authorities are preparing to tighten the rules for digital asset operators. They intend to raise the security standards for these entities to the same level as traditional financial companies. This shift includes adding obligations to digital asset laws to ensure information technology (IT) safety and establishing liability for damages caused by security failures.

These regulations may allow for the preemptive blocking of deposits and withdrawals if a hacking attempt is suspected. The goal is to prevent the rapid movement of assets during an active breach. These measures reflect a broader trend of treating digital assets with the same level of scrutiny as traditional banking systems to protect the wider economy.

Private sector AI security tools

Private companies are also racing to develop specialized tools to manage the risks of generative AI. LloydK recently launched DO-GATE, a security gateway designed to help enterprises use generative AI safely. Meanwhile, Saltware secured a patent for its AI security solution, SafiGuardian, which focuses on detecting sensitive information within AI systems. These tools are designed to prevent employees from inadvertently sharing private data with public AI models.

Other tech firms are expanding their security offerings to capture the growing corporate market. Polaris Office is running a free corporate AD security assessment campaign in partnership with Polaris AI Handy. Additionally, Lotte Innovate has teamed up with Thales to expand its AI security business and unveil premium security monitoring based on artificial intelligence.

The demand for these services is being driven by the fact that AI is moving from simple task support to core IT system development. Companies like BNK Financial Group are already moving to introduce AI-based vulnerability management and 24-hour integrated monitoring. These private initiatives complement the government's efforts by providing the specific technical layers needed to secure corporate data.

New threats and automated defense

The nature of cyberattacks is changing as hackers use AI to find vulnerabilities faster than humans can patch them. For instance, attackers are increasingly using AI-based penetration testing platforms like Antifox to steal authentication information from cloud environments like Amazon Web Services (AWS). This shift is forcing companies to move away from manpower-centered response systems toward automated, AI-driven defenses.

A new type of threat called "LLMjacking" is also emerging, where hackers steal computing power to run large language models. Unlike traditional cryptojacking, which siphons resources slowly, LLMjacking can generate massive usage fees in a very short window. Some companies have reported fraudulent usage fees exceeding $100,000 in a single day, with cumulative damages reaching $1 million before the access was blocked.

To counter these threats, firms are developing automated scripts for vulnerability diagnosis. Softfreak plans to launch a solution called F-Avata Shell this year to automatically generate these scripts using AI. By using AI to defend against AI, these companies hope to keep pace with the speed of modern cybercrime and protect the underlying infrastructure of the digital economy.