The government rejected a 9.1 billion won ($6.78 million) budget request for an understaffed AI-powered cybersecurity center, raising concerns over reliance on OpenAI and widening security gaps.
A projection of cyber code on a hooded figure is seen in this picture taken on May 13, 2017.REUTERS/YONHAP
Korea launched an AI-powered cyber defense center in June to use top-tier AI models to identify security vulnerabilities in national infrastructure and other key systems. But the Ministry of Planning and Budget declined to include the funding by claiming that support should “prioritize small and midsize companies with weaker cybersecurity capabilities.”
The Korea Internet & Security Agency (KISA) requested 9.1 billion won ($6.78 million) for next year to operate its center dedicated to managing cybersecurity vulnerabilities, but the entire budget request was rejected, based on data the Ministry of Science and ICT submitted to Rep. Lee Ju-hee of the ruling Democratic Party.
The funding cutoff is raising concerns that the project could falter at a time when the center is already short-staffed.
The center was originally planned to have at least 40 employees, including private-sector experts. It currently operates with just 24 KISA employees who were reassigned to the unit.
The concern comes as several major financial companies, such as Hana Bank, KB Kookmin Bank and Shinhan Bank, suffered back-to-back leaks of customer information in recent hacking attacks.
ATMs are seen in Seoul on June 28.NEWS1
Its target for this year was to inspect 50 types of commercial software used in infrastructure and equipment, as well as 200 types of open-source software.
In reality, the center has inspected only 24 types of commercial software and 144 types of open-source software over the past four months.
The lack of funding for AI model licenses is also raising concerns over a potential security gap.
A domestically developed AI model specialized for cybersecurity is not expected to be completed until July 2027. Until then, the center relies entirely on OpenAI’s GPT-5.5-Cyber model.
The center was able to detect vulnerabilities this year because OpenAI provided the cybersecurity-focused AI model free of charge.
OpenAI logoREUTERS/YONHAP
“The center will have no choice but to rely entirely on the AI model OpenAI provides for free if it does not receive a dedicated budget,” an ICT Ministry official said. “That could leave Korea’s national cybersecurity system increasingly dependent on OpenAI.”
Experts warn that the funding gap could leave Korea exposed as hackers race ahead with more advanced AI tools.
“Hacker groups are already using frontier-level AI models for attacks, so we cannot simply sit idle until development is completed,” Prof. Youm Heung-youl of information security at Soonchunhyang University said.
This article was originally written in Korean and translated by a bilingual reporter with the help of generative AI tools. It was then edited by a native English-speaking editor. All AI-assisted translations are reviewed and refined by our newsroom.
