OpenAI partners with South Korea's Ministry of Science and ICT to counter AI hacker threats

South Korea's Ministry of Science and ICT is joining forces with OpenAI, the developer of ChatGPT, to bolster its defenses against artificial intelligence-driven cybersecurity threats. This partnership comes in response to the growing global concern over AI security vulnerabilities, highlighted by the recent emergence of models like Anthropic's 'Claude Mythos' preview.

The ministry announced its official participation in OpenAI's "Trustworthy AI Program" (TAC) on May 27. Launched in February, TAC provides select government and institutional entities with access to OpenAI's cutting-edge models for detecting, analyzing, and defending against cyberattacks. South Korea becomes the third nation, after the United States and Canada, and alongside Japan, to join this program at a national level, with the Korea Internet & Security Agency (KISA) managing the practical aspects.

“

The latest cyber AI capabilities should not be confined to a select few but should be utilized by South Korea's key defense entities to strengthen collective security and public safety.

Through TAC, the South Korean government will gain access to OpenAI's specialized 'GPT-5.5 Cyber Model.' This model is an advancement over the GPT-5.4 Cyber Model, which OpenAI released shortly after Anthropic unveiled its 'Claude Mythos Preview' capable of easily identifying critical cybersecurity flaws. The GPT-5.5 Cyber Model has reportedly shown performance slightly exceeding that of the Mythos Preview, according to analyses by the UK's AI Safety Institute.

OpenAI plans to extend access to these models beyond government bodies to include relevant public institutions and key domestic companies involved in national core industries. Jason Kwon, OpenAI's Chief Strategy Officer, emphasized the importance of sharing advanced AI capabilities, stating, "The latest cyber AI capabilities should not be confined to a select few but should be utilized by South Korea's key defense entities to strengthen collective security and public safety." The ministry and OpenAI intend to continue discussions on leveraging AI models for cybersecurity, building on this TAC program participation.

“

We will cooperate closely with the government, public institutions, and companies to strengthen South Korea's capabilities in the cybersecurity field.