Prompt injection: research

Journal articles, academic blogs, think-tank reports and papers in this section, newest first. Back to the news.

arXiv (research papers) Ravikumar Balakrishnan SI

Reading Between the Pixels: Linking Text-Image Embedding Alignment to Typographic Attack Success on Vision-Language Models

We study typographic prompt injection attacks on vision-language models (VLMs), where adversarial text is rendered as images to bypass safety mechanisms, posing a growing threat as VLMs serve as the perceptual backbone of autonomous agents, from browser automation and computer-use systems to camera-equipped embodied agents. In practice, the attack surface is heterogeneous: adversarial text appears at varying font sizes and under diverse visual conditions, while the growing…

Read at arXiv (research papers)

arXiv (research papers) Yulin Chen SI

WebAgentGuard: A Reasoning-Driven Guard Model for Detecting Prompt Injection Attacks in Web Agents

Web agents powered by vision-language models (VLMs) enable autonomous interaction with web environments by perceiving and acting on both visual and textual webpage content to accomplish user-specified tasks. However, they are highly vulnerable to prompt injection attacks, where adversarial instructions embedded in HTML or rendered screenshots can manipulate agent behavior and lead to harmful outcomes such as information leakage. Existing defenses, including system prompt…

Read at arXiv (research papers)

arXiv (research papers) Qingchao Shen SI

TEMPLATEFUZZ: Fine-Grained Chat Template Fuzzing for Jailbreaking and Red Teaming LLMs

Large Language Models (LLMs) are increasingly deployed across diverse domains, yet their vulnerability to jailbreak attacks, where adversarial inputs bypass safety mechanisms to elicit harmful outputs, poses significant security risks. While prior work has primarily focused on prompt injection attacks, these approaches often require resource-intensive prompt engineering and overlook other critical components, such as chat templates. This paper introduces TEMPLATEFUZZ, a…

Read at arXiv (research papers)

arXiv (research papers) Anes Abdennebi SI

Fully Homomorphic Encryption on Llama 3 model for privacy preserving LLM inference

The applications of Generative Artificial Intelligence (GenAI) and their intersections with data-driven fields, such as healthcare, finance, transportation, and information security, have led to significant improvements in service efficiency and low latency. However, this synergy raises serious concerns regarding the security of large language models (LLMs) and their potential impact on the privacy of companies and users' data. Many technology companies that incorporate LLMs…

Read at arXiv (research papers)

arXiv (research papers) Adam Stein SI

Detecting Safety Violations Across Many Agent Traces

To identify safety violations, auditors often search over large sets of agent traces. This search is difficult because failures are often rare, complex, and sometimes even adversarially hidden and only detectable when multiple traces are analyzed together. These challenges arise in diverse settings such as misuse campaigns, covert sabotage, reward hacking, and prompt injection. Existing approaches struggle here for several reasons. Per-trace judges miss failures that only…

Read at arXiv (research papers)

arXiv (research papers) Wei Zhao SI

ClawGuard: A Runtime Security Framework for Tool-Augmented LLM Agents Against Indirect Prompt Injection

Tool-augmented Large Language Model (LLM) agents have demonstrated impressive capabilities in automating complex, multi-step real-world tasks, yet remain vulnerable to indirect prompt injection. Adversaries exploit this weakness by embedding malicious instructions within tool-returned content, which agents directly incorporate into their conversation history as trusted observations. To address these vulnerabilities, we introduce \textsc{ClawGuard}, a novel runtime security…

Read at arXiv (research papers)

arXiv (research papers) Xuwei Ding SI

The Blind Spot of Agent Safety: How Benign User Instructions Expose Critical Vulnerabilities in Computer-Use Agents

Computer-use agents (CUAs) can now autonomously complete complex tasks in real digital environments, but when misled, they can also be used to automate harmful actions programmatically. Existing safety evaluations largely target explicit threats such as misuse and prompt injection, but overlook a subtle yet critical setting where user instructions are entirely benign and harm arises from the task context or execution outcome. We introduce OS-BLIND, a benchmark that evaluates…

Read at arXiv (research papers)

arXiv (research papers) Guijia Zhang SI

STARS: Skill-Triggered Audit for Request-Conditioned Invocation Safety in Agent Systems

Autonomous language-model agents increasingly rely on installable skills and tools to complete user tasks. Static skill auditing can expose capability surface before deployment, but it cannot determine whether a particular invocation is unsafe under the current user request and runtime context. We therefore study skill invocation auditing as a continuous-risk estimation problem: given a user request, candidate skill, and runtime context, predict a score that supports ranking…

Read at arXiv (research papers)

arXiv (research papers) Guangyu Gong SI

PlanGuard: Defending Agents against Indirect Prompt Injection via Planning-based Consistency Verification

Large Language Model (LLM) agents are increasingly integrated into critical systems, leveraging external tools to interact with the real world. However, this capability exposes them to Indirect Prompt Injection (IPI), where attackers embed malicious instructions into retrieved content to manipulate the agent into executing unauthorized or unintended actions. Existing defenses predominantly focus on the pre-processing stage, neglecting the monitoring of the model's actual…

Read at arXiv (research papers)

arXiv (research papers) Guiyao Tie SI

BadSkill: Backdoor Attacks on Agent Skills via Model-in-Skill Poisoning

Agent ecosystems increasingly rely on installable skills to extend functionality, and some skills bundle learned model artifacts as part of their execution logic. This creates a supply-chain risk that is not captured by prompt injection or ordinary plugin misuse: a third-party skill may appear benign while concealing malicious behavior inside its bundled model. We present BadSkill, a backdoor attack formulation that targets this model-in-skill threat surface. In BadSkill, an…

Read at arXiv (research papers)

arXiv (research papers) Zedian Shao SI

Leave My Images Alone: Preventing Multi-Modal Large Language Models from Analyzing Images via Visual Prompt Injection

Multi-modal large language models (MLLMs) have emerged as powerful tools for analyzing Internet-scale image data, offering significant benefits but also raising critical safety and societal concerns. In particular, open-weight MLLMs may be misused to extract sensitive information from personal images at scale, such as identities, locations, or other private details. In this work, we propose ImageProtector, a user-side method that proactively protects images before sharing by…

Read at arXiv (research papers)

arXiv (research papers) Runpeng Geng SI

PIArena: A Platform for Prompt Injection Evaluation

Prompt injection attacks pose serious security risks across a wide range of real-world applications. While receiving increasing attention, the community faces a critical gap: the lack of a unified platform for prompt injection evaluation. This makes it challenging to reliably compare defenses, understand their true robustness under diverse attacks, or assess how well they generalize across tasks and benchmarks. For instance, many defenses initially reported as effective were…

Read at arXiv (research papers)

arXiv (research papers) Nicolás E. Díaz Ferreyra SI

Security Concerns in Generative AI Coding Assistants: Insights from Online Discussions on GitHub Copilot

Generative Artificial Intelligence (GenAI) has become a central component of many development tools (e.g., GitHub Copilot) that support software practitioners across multiple programming tasks, including code completion, documentation, and bug detection. However, current research has identified significant limitations and open issues in GenAI, including reliability, non-determinism, bias, and copyright infringement. While prior work has primarily focused on assessing the…

Read at arXiv (research papers)

arXiv (research papers) Wenkui Yang SI

Are GUI Agents Focused Enough? Automated Distraction via Semantic-level UI Element Injection

Existing red-teaming studies on GUI agents face two fundamental limitations: adversarial perturbations require white-box access unavailable in commercial deployments, while prompt injection is increasingly neutralized by stronger safety alignment. To study robustness under a more practical threat model, we propose Semantic-level UI Element Injection, a black-box red-teaming paradigm that overlays safety-aligned and harmless UI elements onto screenshots to misdirect the…

Read at arXiv (research papers)

arXiv (research papers) Hengkai Ye SI

TRUSTDESC: Preventing Tool Poisoning in LLM Applications via Trusted Description Generation

Large language models (LLMs) increasingly rely on external tools to perform time-sensitive tasks and real-world actions. While tool integration expands LLM capabilities, it also introduces a new prompt-injection attack surface: tool poisoning attacks (TPAs). Attackers manipulate tool descriptions by embedding malicious instructions (explicit TPAs) or misleading claims (implicit TPAs) to influence model behavior and tool selection. Existing defenses mainly detect anomalous…

Read at arXiv (research papers)

arXiv (research papers) Yen-Shan Chen SI

TraceSafe: A Systematic Assessment of LLM Guardrails on Multi-Step Tool-Calling Trajectories

As large language models (LLMs) evolve from static chatbots into autonomous agents, the primary vulnerability surface shifts from final outputs to intermediate execution traces. While safety guardrails are well-benchmarked for natural language responses, their efficacy remains largely unexplored within multi-step tool-use trajectories. To address this gap, we introduce **TraceSafe-Bench**, the first comprehensive benchmark specifically designed to assess mid-trajectory…

Read at arXiv (research papers)

arXiv (research papers) Manish Bhatt SI

The Defense Trilemma: Why Prompt Injection Defense Wrappers Fail?

We prove that no continuous, utility-preserving wrapper defense-a function $D: X\to X$ that preprocesses inputs before the model sees them-can make all outputs strictly safe for a language model with connected prompt space, and we characterize exactly where every such defense must fail. We establish three results under successively stronger hypotheses: boundary fixation-the defense must leave some threshold-level inputs unchanged; an $ε$-robust constraint-under Lipschitz…

Read at arXiv (research papers)

arXiv (research papers) Purva Chiniya SI

Gradient-Controlled Decoding: A Safety Guardrail for LLMs with Dual-Anchor Steering

Large language models (LLMs) remain susceptible to jailbreak and direct prompt-injection attacks, yet the strongest defensive filters frequently over-refuse benign queries and degrade user experience. Previous work on jailbreak and prompt injection detection such as GradSafe, detects unsafe prompts with a single "accept all" anchor token, but its threshold is brittle and it offers no deterministic guarantee that harmful content will not be emitted once decoding begins. We…

Read at arXiv (research papers)

arXiv (research papers) Geert Trooskens XY.AI Labs, Palo Alto, CA SI

Compiled AI: Deterministic Code Generation for LLM-Based Workflow Automation

We study compiled AI, a paradigm in which large language models generate executable code artifacts during a compilation phase, after which workflows execute deterministically without further model invocation. This paradigm has antecedents in prior work on declarative pipeline optimization (DSPy) and hybrid neural-symbolic planning (LLM+P); our contribution is a systems-oriented study of its application to high-stakes enterprise workflows, with particular emphasis on…

Read at arXiv (research papers)

arXiv (research papers) Guy Azov SI

SALLIE: Generation-Free Hidden-State Detection of Jailbreaks and Prompt Injections Across Text and Vision

Large Language Models (LLMs) and Vision-Language Models (VLMs) are vulnerable to jailbreaks and prompt injections delivered through text or images. Existing defenses often narrow threat coverage or add inference cost through input transformations, repeated model queries, generated responses, or threat-specific detectors. We present SALLIE (Single-Pass Activation Lookup for Layerwise Input Evaluation), a white-box, generation-free detector with a common architecture for…

Read at arXiv (research papers)

arXiv (research papers) Zhuowen Yuan SI

ShieldNet: Network-Level Guardrails against Emerging Supply-Chain Injections in Agentic Systems

Existing research on LLM agent security mainly focuses on prompt injection and unsafe input/output behaviors. However, as agents increasingly rely on third-party tools and MCP servers, a new class of supply-chain threats has emerged, where malicious behaviors are embedded in seemingly benign tools, silently hijacking agent execution, leaking sensitive data, or triggering unauthorized actions. Despite their growing impact, there is currently no comprehensive benchmark for…

Read at arXiv (research papers)

arXiv (research papers) Fariha Tanjim Shifat SI

LLM-Enabled Open-Source Systems in the Wild: An Empirical Study of Vulnerabilities in GitHub Security Advisories

Large language models (LLMs) are increasingly embedded in open-source software (OSS) ecosystems, creating complex interactions among natural language prompts, probabilistic model outputs, and execution-capable components. However, it remains unclear whether traditional vulnerability disclosure frameworks adequately capture these model-mediated risks. To investigate this, we analyze 295 GitHub Security Advisories published between January 2025 and January 2026 that reference…

Read at arXiv (research papers)

arXiv (research papers) Dalal Alharthi SI

Automating Cloud Security and Forensics Through a Secure-by-Design Generative AI Framework

As cloud environments become increasingly complex, cybersecurity and forensic investigations must evolve to meet emerging threats. Large Language Models (LLMs) have shown promise in automating log analysis and reasoning tasks, yet they remain vulnerable to prompt injection attacks and lack forensic rigor. To address these dual challenges, we propose a unified, secure-by-design GenAI framework that integrates PromptShield and the Cloud Investigation Automation Framework…

Read at arXiv (research papers)

arXiv (research papers) Wenhui Zhu SI

Your Agent is More Brittle Than You Think: Uncovering Indirect Injection Vulnerabilities in Agentic LLMs

The rapid deployment of open-source frameworks has significantly advanced the development of modern multi-agent systems. However, expanded action spaces, including uncontrolled privilege exposure and hidden inter-system interactions, pose severe security challenges. Specifically, Indirect Prompt Injections (IPI), which conceal malicious instructions within third-party content, can trigger unauthorized actions such as data exfiltration during normal operations. While current…

Read at arXiv (research papers)

arXiv (research papers) Jackson Wang SI

AttackEval: A Systematic Empirical Study of Prompt Injection Attack Effectiveness Against Large Language Models

Prompt injection has emerged as a critical vulnerability in large language model (LLM) deployments, yet existing research is heavily weighted toward defenses. The attack side -- specifically, which injection strategies are most effective and why -- remains insufficiently studied.We address this gap with AttackEval, a systematic empirical study of prompt injection attack effectiveness. We construct a taxonomy of ten attack categories organized into three parent groups…

Read at arXiv (research papers)

arXiv (research papers) Yilin Xiao SI

LogicPoison: Logical Attacks on Graph Retrieval-Augmented Generation

Graph-based Retrieval-Augmented Generation (GraphRAG) enhances the reasoning capabilities of Large Language Models (LLMs) by grounding their responses in structured knowledge graphs. Leveraging community detection and relation filtering techniques, GraphRAG systems demonstrate inherent resistance to traditional RAG attacks, such as text poisoning and prompt injection. However, in this paper, we find that the security of GraphRAG systems fundamentally relies on the…

Read at arXiv (research papers)

arXiv (research papers) Bowen Wei SI

ClawSafety: "Safe" LLMs, Unsafe Agents

Personal AI agents like OpenClaw run with elevated privileges on users' local machines, where a single successful prompt injection can leak credentials, redirect financial transactions, or destroy files. This threat goes well beyond conventional text-level jailbreaks, yet existing safety evaluations fall short: most test models in isolated chat settings, rely on synthetic environments, and do not account for how the agent framework itself shapes safety outcomes. We introduce…

Read at arXiv (research papers)

arXiv (research papers) Yanting Wang SI

AgentWatcher: A Rule-based Prompt Injection Monitor

Large language models (LLMs) and their applications, such as agents, are highly vulnerable to prompt injection attacks. State-of-the-art prompt injection detection methods have the following limitations: (1) their effectiveness degrades significantly as context length increases, and (2) they lack explicit rules that define what constitutes prompt injection, causing detection decisions to be implicit, opaque, and difficult to reason about. In this work, we propose…

Read at arXiv (research papers)

arXiv (research papers) Cormac Guerin SI

KAIJU: An Executive Kernel for Intent-Gated Execution of LLM Agents

Tool-calling autonomous agents based on large language models using ReAct exhibit three limitations: serial latency, quadratic context growth, and vulnerability to prompt injection and hallucination. Recent work moves towards separating planning from execution but in each case the model remains coupled to the execution mechanics. We introduce a system-level abstraction for LLM agents which decouples the execution of agent workflows from the LLM reasoning layer. We define two…

Read at arXiv (research papers)

arXiv (research papers) Chong Xiang SI

Architecting Secure AI Agents: Perspectives on System-Level Defenses Against Indirect Prompt Injection Attacks

AI agents, predominantly powered by large language models (LLMs), are vulnerable to indirect prompt injection, in which malicious instructions embedded in untrusted data can trigger dangerous agent actions. This position paper discusses our vision for system-level defenses against indirect prompt injection attacks. We articulate three positions: (1) dynamic replanning and security policy updates are often necessary for dynamic tasks and realistic environments; (2) certain…

Read at arXiv (research papers)

Older stories