Red Teaming the Mind of the Machine: A Systematic Evaluation of Prompt Injection and Jailbreak Vulnerabilities in LLMs
Fuente:
arXiv
Saved in:
| Main Author: | |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908361558065152 |
|---|---|
| author | Pathade, Chetan |
| author_facet | Pathade, Chetan |
| contents | Large Language Models (LLMs) are increasingly integrated into consumer and enterprise applications. Despite their capabilities, they remain susceptible to adversarial attacks such as prompt injection and jailbreaks that override alignment safeguards. This paper provides a systematic investigation of jailbreak strategies against various state-of-the-art LLMs. We categorize over 1,400 adversarial prompts, analyze their success against GPT-4, Claude 2, Mistral 7B, and Vicuna, and examine their generalizability and construction logic. We further propose layered mitigation strategies and recommend a hybrid red-teaming and sandboxing approach for robust LLM security. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_04806 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Red Teaming the Mind of the Machine: A Systematic Evaluation of Prompt Injection and Jailbreak Vulnerabilities in LLMs Pathade, Chetan Cryptography and Security Computation and Language Large Language Models (LLMs) are increasingly integrated into consumer and enterprise applications. Despite their capabilities, they remain susceptible to adversarial attacks such as prompt injection and jailbreaks that override alignment safeguards. This paper provides a systematic investigation of jailbreak strategies against various state-of-the-art LLMs. We categorize over 1,400 adversarial prompts, analyze their success against GPT-4, Claude 2, Mistral 7B, and Vicuna, and examine their generalizability and construction logic. We further propose layered mitigation strategies and recommend a hybrid red-teaming and sandboxing approach for robust LLM security. |
| title | Red Teaming the Mind of the Machine: A Systematic Evaluation of Prompt Injection and Jailbreak Vulnerabilities in LLMs |
| topic | Cryptography and Security Computation and Language |
| url | https://arxiv.org/abs/2505.04806 |