This article explores how large language models (LLMs) can be used for both defensive and offensive purposes in cybersecurity, highlighting the rise of malicious models like WormGPT and WormGPT 4. These tools bypass ethical constraints, making cybercrime more accessible for less skilled attackers. The piece details their capabilities, including generating phishing content and malware, and discusses the implications for the threat landscape.

An underground AI tool called SpamGPT is emerging as a CRM for cybercriminals, providing advanced marketing capabilities that enable more effective and targeted spam campaigns. This tool is designed to streamline operations for cybercriminals, offering features similar to legitimate business software, thus enhancing their ability to execute scams and phishing attacks. The rise of such tools highlights the ongoing challenges in cybersecurity and the increasing sophistication of cybercriminal activities.

AI is transforming the cybercrime landscape by enhancing existing attack methods rather than creating new threats, making cybercriminal activities more efficient and accessible. The panel at RSA Conference 2025 emphasized the importance of adapting defense strategies to counter AI-driven attacks, highlighting the need for international cooperation and innovative security frameworks. As AI continues to evolve, both defenders and threat actors will need to adapt rapidly to the changing dynamics of cyber threats.