Measuring Real-World Prompt Injection Attacks in LLM-based Resume Screening

Mohan Zhang, The University of North Carolina at Chapel Hill; Yuqi Jia, Duke University; Zhen Tan, Arizona State University; Steven Jiang, hireEZ; Neil Zhenqiang Gong, Duke University; Tianlong Chen, The University of North Carolina at Chapel Hill; Dawn Song, UC Berkeley

LLMs are vulnerable to prompt injection attacks. However, this vulnerability has been primarily demonstrated conceptually in academic studies or through a few anecdotal case studies. Its prevalence and impact in real-world LLM-based applications are largely unexplored. In this work, we present the first systematic study of prompt-injection attacks in a widely used application: LLM-based resume screening. Our analysis is based on approximately 200K real-world resumes collected over multiple years by AnonyCom (The company's name is intentionally anonymized for double-blind review purposes). We first design tailored methods to detect prompt injection in resumes. Manual verification on a small-scale dataset demonstrates that our detectors achieve high precision and outperform state-of-the-art general-purpose detectors. We then apply our detector to the full resume dataset and conduct a comprehensive measurement study of real-world prompt injection attacks. Our analysis reveals several intriguing findings: approximately 1% of resumes contain hidden prompt injections; the prevalence of such injected resumes has increased noticeably over the past one to two years; and more than 90% of injected prompts do not use explicit instructions. These results provide the first evidence of large-scale prompt injection in real-world LLM-based applications and lay the groundwork for future studies to understand and mitigate such attacks.

USENIX Security '26 Open Access Sponsored by
King Abdullah University of Science and Technology (KAUST)

Open Access Media

USENIX is committed to Open Access to the research presented at our events. Papers and proceedings are freely available to everyone once the event begins. Any video, audio, and/or slides that are posted after the event are also free and open to everyone. Support USENIX and our commitment to Open Access.

BibTeX
@inproceedings {320739,
author = {Mohan Zhang and Yuqi Jia and Zhen Tan and Steven Jiang and Neil Zhenqiang Gong and Tianlong Chen and Dawn Song},
title = {Measuring {Real-World} Prompt Injection Attacks in {LLM-based} Resume Screening},
booktitle = {35th USENIX Security Symposium (USENIX Security 26)},
year = {2026},
isbn = {978-1-939133-58-8},
address = {Baltimore, MD},
pages = {1427--1445},
url = {https://www.usenix.org/conference/usenixsecurity26/presentation/zhang-mohan},
publisher = {USENIX Association},
month = aug
}