AI-Based CYBERSECURITY: Benefits and Limitations

Purpose: This study intends to examine the efficiency and limitations of AI-based cyber defense systems by applying game theory and to explore the direction of the development of an AI-based national cyber defense system. In cybersecurity attackers and defenders can choose strategies and make a decision based on their resources, to attain the rewards, while anticipating the actions from opposing players. Method: For better analysis, this article use Game theory an analytical framework to identify the position between the attacker and defender in the Cyber domain. Game theory has been used to observe competition among multiple competitors(players) fighting under pre-set rules. Game theory is appropriate for analyzing cyber interaction between defenders, attackers, and users and what happens as a result. Results: To respond to infringement incidents, first, for artificial intelligence to more efficiently detect and respond quickly to security threats, it is necessary to analyze vast amounts of security data with human expe-rience and knowledge and enter accurate data for artificial intelligence to learn. Second, it is necessary to strengthen the self-learning ability of security solutions to apply signature analysis, behavior analysis technolo-gy, and machine learning technology to enable automated detection and response to AI-based cyber attacks. Conclusion: Although there are many limitations and risks of artificial intelligence, it is important to concern about how to use artificial intelligence usefully for the welfare and prosperity of mankind. First, cybersecurity is directly related to national security, and the government has to enhance an AI-based cybersecurity system. Second, to cope with increasingly diverse and evolving external cyber-attacks. Third, it is necessary to develop R & D investment and professional human resources to build an AI-based cybersecurity system. Fourth, it is re-quired to overhaul related legal systems to strengthen AI-based cybersecurity. Although there are many limita-tions and risks of artificial intelligence, it is important to concern about how to use artificial intelligence usefully for the welfare and prosperity of mankind.