The Presidential Security Service (PSS) of South Korea has officially introduced a generative artificial intelligence (AI) system, marking the beginning of an 'AI security era'. On the 25th, the PSS released a statement indicating that this AI system was activated in mid-February to proactively respond to the rapidly changing security landscape and to maximize operational efficiency. The implementation is intended to establish a data-driven scientific security framework aiming for 'perfect security'. This innovative system incorporates a 'Retrieval-Augmented Generation' (RAG) technology that has been trained on decades of internal data accumulated by the PSS. This technology is designed to provide accurate responses by sourcing relevant information, effectively addressing known limitations of generative AI, such as 'hallucination', where AI generates incorrect or fictional responses as if they were factual. It is aimed at increasing the accuracy and reliability of information provided by the AI. Unique to this initiative is the development of a closed network that is completely isolated from external internet access to ensure security; this measure prevents potential information breaches while still enabling the use of generative AI functionalities in their operations. The PSS staff will now utilize this chatbot-like system for diverse tasks including event risk assessment, legal research, and document translation and summarization. The PSS plans to expand the system's accuracy and application scope based on accumulated data, ultimately aiming to develop an 'intelligent security platform' capable of countering emerging threats that may misuse AI technologies.