A Review of LLM Security: Threats and Mitigations

Date

2025-1-23

Author

Günay, Bengi

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

225
views

182
downloads

Cite This

Large Language Models (LLMs) are advanced AI systems trained on massive datasets to process and generate natural language. These models can understand, summarize, and create human-like text with remarkable accuracy. Since the release of ChatGPT, LLMs have gained widespread popularity worldwide. While they offer numerous benefits, they also present significant security challenges. This study explores the security dimensions of LLM technology. It categorizes and examines potential threats in detail and discusses effective mitigation techniques to address these vulnerabilities. Moreover, as part of this study, we will examine and analyze case studies of LLM vulnerabilities from PortSwigger’s Web Security Academy.

Subject Keywords

Large language models, attacks, mitigations, threats, security, vulnerability

URI

https://hdl.handle.net/11511/113064

Collections

Graduate School of Informatics, Term Project