CORTEXA
← Browse
arxivcs.AI2026-07-03

CrackedPDFs: A Controlled Benchmark for Hidden Prompt Injection in PDFs

Pukaphol Thienpreecha

Document-based LLM systems often flatten a PDF before guardrails inspect it. That step can discard evidence that an instruction was never visible to the user. We introduce CrackedPDFs, a controlled benchmark for hidden prompt injection in PDFs. The benchmark contains 29,322 generated PDFs from 4,983 base docu ments. It includes 9,774 injected files and 19,548 benign or matched-confounder files. We evaluate PromptGuard and a rule baseline. We also evaluate structural only learned models and a sanitized hybrid detector. The evaluation uses held-out provenance splits and paired benign-confounder controls. It also uses label-shuffle checks and shortcut audits. On a 2,919-document held-out test set, the hybrid de tector reaches 0.960 F1. ROC-AUC is 0.998 and PR-AUC is 0.997. It also ranks injected files above matched benign confounders in 95.9% of 973 pairs. Prompt Guard has low recall when given extracted text only. Structural-only learned mod els are weak under paired controls. A text-only TF-IDF model reaches perfect held-out scores but fails shortcut audits. These results show that document-aware hybrid detection is useful under controlled paired evaluation. They do not show broad real-world robustness or reliable cross-family generalization.

View free PDFSource page

Related papers

arxivcs.CRcs.AI2026-06-29

Security--Fidelity Tradeoffs: The Hidden Cost of Prompt Injection Defense

Mitchell Hermon, Rahul Gupta, Weitong Ruan, Ekraam Sabir, Haohan Wang

We identify a security-fidelity tradeoff in defending LLMs against indirect prompt injection: defenses resist injected instructions largely by suppressing untrusted text, which corrupts tasks that must preserve it, such as translation and document editing. Attack-success metrics…

View free PDFSource page
arxivcs.CRcs.AI2026-07-21

CPInj: Uncovering Prompt Injection Risks in Textual Collaborative Prompt Optimization

Xinting Liao, Behnoosh Zamanlooy, Masoumeh Shafieinejad, David B. Emerson, Ruinan Jin, Deval Pandya, et al.

Textual Collaborative Prompt Optimization (TCPO) extends Textgrad (Yuksekgonul et al., 2025) to a decentralized setting by allowing multiple clients to jointly improve prompts for large language models (LLMs) while keeping their data locally. Its reliance on free-form textual upd…

View free PDFSource page
arxivcs.AI2026-06-25

Prompt Injection in Automated Résumé Screening with Large Language Models: Single and Multi-Injection Settings

Preet Baxi, Jiannan Xu, Jane Yi Jiang, Stefanus Jasin

Large language models (LLMs) are increasingly used to screen and rank job applicants, creating incentives for candidates to strategically manipulate algorithmic hiring systems. We study prompt injection in automated résumé screening, defined as subtle self-promotional text that i…

View free PDFSource page
arxivcs.CRcs.AIcs.MA2026-07-16

Bad Memory: Evaluating Prompt Injection Risks from Memory in Agentic Systems

Soham Gadgil, David Alexander, Sai Sunku, Franziska Roesner

A growing class of agentic systems maintain persistent state across sessions through memory files, behavioral preferences, and knowledge bases. While this makes agents more useful and self-improving, it also creates a new attack surface for prompt injections in which malicious in…

View free PDFSource page
arxivcs.SEcs.AI2026-07-16

MM-IssueLoc: A Controlled Benchmark for Evaluating Visual Evidence in Multimodal Repository-Level Issue Localization

Shaoxiong Zhan, Shi Hu, Boyu Feng, Hai Lin, Andrew Gong, Zhengda Zhou, et al.

Real repository issues routinely include visual evidence such as screenshots, error dialogs, rendered UI states, and logs, yet repository-level issue localization is evaluated mostly as a text-only task. Existing multimodal SE benchmarks evaluate end-to-end repair, entangling loc…

View free PDFSource page