South Korea’s Ministry of Science and ICT said it held an AI security hands-on workshop with OpenAI on May 18. The foreign ministry, the National Intelligence Service, the Financial Services Commission, the National AI Strategy Committee, the AI Safety Research Institute, the Korea Internet & Security Agency (KISA) and the Financial Security Institute also took part.

At the workshop, Sasha Baker (사샤 베이커), OpenAI’s head of national security policy, explained its Trusted Access for Cyber (TAC) programme to South Korean government officials. She also demonstrated cybersecurity-related functions of the latest AI models. The two sides also discussed ways to cooperate in AI security.

The ministry asked OpenAI to share information and cooperate in responding to AI security threats. The two sides agreed to continue working-level discussions on the use of AI models for cybersecurity, using the meeting as a starting point.

Baker said close public-private cooperation is essential to respond effectively to cyber threats in the AI era. She said AI can be used safely and cyber resilience strengthened when governments, companies, infrastructure operators, research institutions and AI companies cooperate based on a shared sense of responsibility.

Choi Woo-hyuk (최우혁), director general for information security and network policy at the ministry, said there was a shared understanding that AI should be a tool for defenders, not attackers, and that close cooperation across the public and private sectors is needed to respond to AI security threats. He said the ministry would strengthen South Korea’s AI security capabilities through active cooperation and working-level discussions with global AI companies.