Exploring the Dual Role of LLMs in Cybersecurity: Threats and Defenses
摘要
Large Language Models (LLMs) pose risks for cybersecurity since they facilitate minimal cost creation of malware, phishing messages, and malicious chatbots. At the same time, LLMs can help defend against cyberattacks. This chapter reviews security research around the risks and benefits of LLMs.