Logo
International Journal of
Advanced Education and Research

Search

ARCHIVES
VOL. 10, ISSUE 2 (2025)
Comprehensive analysis of security and privacy challenges in large language models
Authors
Suhani Goyal, Saksham kaunt, Saurabh, Dr. MeenaChaudhary, Dr. NarenderGautam
Abstract

Large Language Models (LLMs) are AI systems that use deep learning to understand and generate human language, performing tasks such as text generation, translation, and question answering. These models work by predicting the next word or sequence based on the input they receive, trained on vast datasets.

Despite their advancements, LLMs face significant security and privacy challenges. Prompt injection attacks manipulate model outputs, while data memorization risks exposing sensitive information from training data. This paper explores these challenges and existing mitigation methods, proposing dynamic prompt filters to counter prompt injection and contextual differential privacy to address data memorization. These solutions aim to enhance both security and privacy, advancing the trustworthiness of LLMs.
Download
Pages:84-88
How to cite this article:
Suhani Goyal, Saksham kaunt, Saurabh, Dr. MeenaChaudhary, Dr. NarenderGautam "Comprehensive analysis of security and privacy challenges in large language models". International Journal of Advanced Education and Research, Vol 10, Issue 2, 2025, Pages 84-88
Download Author Certificate

Please enter the email address corresponding to this article submission to download your certificate.