Navigating the security landscape of large language models in enterprise information systems

Abstract

ABSTRACT In this letter, we present a comprehensive analysis of the security landscape surrounding large language models, leveraging a dataset from 2022 to 2023. We delve into the collaborative and global nature of this field, emphasizing the interdisciplinary approach required to address the evolving challenges. Prominent keywords like ‘language model,’ ‘cybersecurity,’ and ‘data privacy’ underscore central research themes. Beyond analysis, we introduce a cutting-edge deep CNN-based security framework, successfully tested with the KDDCup dataset, resulting in an impressive 94.73% accuracy on the testing dataset. As large language models continue to play a pivotal role in diverse applications, our research underscores the urgency of international cooperation and innovative security measures to ensure their responsible and secure usage in our interconnected world. This letter serves as a foundation for future research, driving the collective effort to safeguard large language models.