Government to Launch Joint Public-Private System for AI Cyber Threats... Domestic Information Security to Transition to Independent AI-Based System Next Year

"Sharing AI Model Vulnerabilities Secured Through International Cooperation"

Baek Gung-hoon, Deputy Prime Minister and Minister of Science and ICT, attended the Science and Technology Ministers' Meeting held at the Government Complex Seoul in Jongno-gu, Seoul on the 30th, speaking on current issues such as the policy direction for data promotion in the era of AI transformation. 2026.4.30 Photo by Jo Yongjun

The government is moving to establish a joint public-private response system to counter evolving artificial intelligence (AI) cyber threats. It will centralize the management of AI vulnerabilities and patch information while strengthening technical support for businesses and government agencies. In addition, starting next year, the domestic information security system will be transitioned to an independent AI technology-based system to prepare for the routine security utilization of high-performance AI.

On May 29, the Ministry of Science and ICT announced its "Private Sector Information Security Promotion Plan to Respond to AI-based Cyber Threats" at the 9th Science and Technology Ministers' Meeting.

Recently, major U.S. big tech companies have launched projects to provide AI models with cybersecurity capabilities at the level of security experts exclusively to select companies. According to the first report of Anthropic's Glasswing Project in May, more than 16,000 vulnerabilities were found in the project participants' software and open-source assets.

As an initial step, the government will establish a joint public-private response system to address the public disclosure of AI vulnerabilities. Centered on the National Security Office of the Blue House, the government will ensure the rapid sharing and dissemination of information on AI vulnerability disclosures, patches, and threat situations, and will set up an emergency system for joint response in the event of security incidents. The Ministry of Science and ICT will operate a central situation room, while each relevant ministry will run its own situation room for the private sector.

A Vulnerability Management Center will be set up within the Korea Internet & Security Agency (KISA) to centralize the management of vulnerabilities and patches, and to promote technical support for relevant ministries and businesses. The KISA Vulnerability Information Portal (KNVD) will be the hub for extensive collection and analysis of vulnerabilities and patches through public disclosure, reports, and sharing with related agencies. This information will be rapidly shared with approximately 28,000 Chief Information Security Officers (CISOs) at companies, private cooperation channels (C-TAS, ISAC), and all government and military situation rooms, forming an emergency response system that swiftly issues action recommendations.

The Ministry of Science and ICT also plans to pilot the use of the latest high-performance AI models—secured through international cooperation—for work related to vulnerabilities and patches, as well as for overall business support.

For major companies that could suffer significant damage from AI security threats, security preparedness will be strengthened. Under the supervision of each responsible ministry, these companies will be required to independently manage asset management, vulnerability inspections, and patch responses, while the government will conduct sector-specific implementation checks.

For small and medium-sized enterprises (SMEs), the government will distribute a web tool that diagnoses current security levels and recommends security investment guidelines and action steps accordingly. It will also support technology for generating and analyzing software bills of materials (SBOMs) to proactively identify and address open-source vulnerabilities that are easily exploited by AI.

The Ministry of Science and ICT plans to provide infrastructure, utilizing high-performance AI models to which it has access, for vulnerability inspections and other support for SME software.

Additionally, to counter AI security threats, the government will continuously monitor domains worldwide and detect and respond to AI-based malicious activities and domains as soon as they are created. In the event of incidents or suspected incidents involving AI services, the "Incident Investigation and Review Committee" will be immediately convened to ensure a swift investigation and prevent the spread of damage.

The Ministry of Science and ICT will begin by securing OpenAI's Government Trust-based Access Program (GTAC) and will continue cooperation to participate in AI security projects with global big tech companies. The ministry will also strengthen collaboration with cybersecurity agencies in allied countries for AI-based threat response.

Looking ahead, the government plans to fully transition the domestic information security system to an independent AI technology-based model from next year, in preparation for the everyday security use and potential weaponization of high-performance AI. Various projects will be planned and executed to establish AI security sovereignty.

Baek Gung-hoon, Deputy Prime Minister and Minister of Science and ICT, stated, "The pace of AI development in the field of cybersecurity is now comparable to that of top-level hackers. If Korea fails to establish a security system and global collaborations suitable for the AI era, its ambitions to become a global top-three AI power could be shaken. Through these measures, we will establish an emergency response system to address large-scale AI-related vulnerability disclosures and will swiftly pursue the establishment of AI security sovereignty based on our own technologies and models."

This content was produced with the assistance of AI translation services.