Prompt injection: research

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

arXiv (research papers) Buzhao Liu SI

Robust Context-Aware Detection of Malicious Instructions in Text

The remarkable instruction-following ability of modern LLMs has enabled their practical use in orchestrating workflows of tool-using agents. This, however, also exposes such agents to attacks which embed malicious instructions in text, such as indirect prompt injection (IPI). One class of approaches to defend against such attacks aims to detect malicious sentences within otherwise benign text. While a number of approaches for this have been proposed, all tend to be brittle…

Read at arXiv (research papers)

arXiv (research papers) Yanting Wang SI

Agent Against Agent: An Agentic System for Automatic Prompt Injection Red Teaming

Prompt injection poses significant security risks to LLM agents. Efficient and effective red-teaming is therefore critical, both for evaluating these risks and for collecting training data to improve defenses. Existing state-of-the-art prompt injection red-teaming methods primarily rely on reinforcement learning (RL), producing attacker models that often generalize poorly to new target LLMs. In this work, we develop PIMiner, an agentic system for prompt injection…

Read at arXiv (research papers)

arXiv (research papers) Longtao Guo SI

LoginTrap: Uncovering Task-Agnostic Phishing-Style Indirect Prompt Injection Attacks against LLM-based Web Agents

LLM-based web agents automate user tasks by observing webpages and executing browser actions on behalf of users. As these agents operate on real web services, login becomes a sensitive authentication boundary because it involves credentials and sensitive information. Existing work shows that malicious webpage content can manipulate web agent actions, but it has not fully examined whether such content can induce login and cause end-to-end private data leakage. We study this…

Read at arXiv (research papers)

arXiv (research papers) Xuebin Li SI

Breadcrumbing Search Agents: Per-Turn Scheming Over Long-Horizon Trajectories

LLM-based search agents are widely used for information-seeking tasks, but their reliance on external tool returns introduces a critical security risk: web content retrieved during execution is untrusted, exposing agents to prompt injection and goal hijacking. Prior work on search-agent safety primarily focuses on static web-content injection, but modern agents issue follow-up queries and cross-check competing sources, so a single injected page is often diluted or rejected…

Read at arXiv (research papers)

arXiv (research papers) Peichun Hua SI

Behavioral Skill Reconstruction: Reconstructing Hidden Functionality from LLM Agent Skills

Closed source agent skills may encode proprietary instructions, scripts, constants, and data. Providers may offer their capabilities as services while keeping the underlying packages hidden. Prior work focuses on prompt injection attacks that directly disclose these artifacts, and existing defenses accordingly aim to prevent such leakage. However, preventing file disclosure does not prevent users from recovering the functionality those files implement. This raises a…

Read at arXiv (research papers)

arXiv (research papers) Shihao Weng SI

AgentAntibody: An Adaptive Immune System for Defending LLM Agents against Prompt Injection

Prompt injection remains a critical threat to LLM agents, yet existing defenses treat each task as a self-contained problem, independent of previous encounters. In practice, user requests are often underspecified: they describe the desired outcome without fully specifying acceptable behavior. An injection can exploit this ambiguity, causing the agent to complete the task in a way the user would reject. As the user's expectations become clearer through concrete cases, a…

Read at arXiv (research papers)

arXiv (research papers) Jia-Chen Zhang SI

Invisible Ink Threats: Adversarial Goals Behind Legitimate Tasks in Computer-Use Agents

Computer-use agents (CUAs), which empower large language models to autonomously operate operating systems and the web, are increasingly vulnerable to indirect prompt injection attacks. A widely adopted defense is the human-in-the-loop paradigm, in which the agent pauses for explicit user confirmation before executing sensitive operations. While effective against conspicuously high-harm attacks, this defense offers little protection against what we term Invisible Ink Threats…

Read at arXiv (research papers)

arXiv (research papers) Qianlong Yang SI

Mitigating Visual Degradation in MLLMs via Spatial-Spectral Visual Anchor Learning

Despite the progress of multimodal large language models (MLLMs), they continue to exhibit deficiencies in visual perception. Following visual instruction tuning, internal MLLM representations rapidly deviate from their original semantic states during inference, causing severe information degradation. While existing methods attempt to leverage external vision foundation models (VFMs) to align internal representations, we find that direct alignment with VFMs enhances visual…

Read at arXiv (research papers)

arXiv (research papers) Dheeraj Mohandas Pai SI

FraudBench: Stress-Testing Policy-Grounded Banking Agents Against Adaptive Fraud

Conversational agents now act for end users through tools while holding access to customer databases and internal policy documents that a caller can reach through dialogue alone. Banking is the clearest case: the same agent that answers a question can also change contact details, reset a PIN, or move money, so ordinary customer service is inseparable from authorization, fraud detection, and policy compliance. Existing financial-fraud benchmarks classify static transactions…

Read at arXiv (research papers)

arXiv (research papers) Fred Zimmerman SI

Copyright Is the Headline; Capability Is the Blind Spot: AI Technology in the Book-Publishing Trade Press, November 2025--August 2026

This rapid evidence review examines 89 articles about artificial intelligence (AI) and book publishing published from November 1, 2025 through August 1, 2026. The purposive corpus spans English-, Chinese-, German-, French-, Spanish-, Portuguese-, Italian-, and Japanese-language publishing coverage; major-newspaper book coverage; and specialist technology commentators. Each item was coded for topic, stance, technical depth, and dominant voice. The press is neither silent nor…

Read at arXiv (research papers)

arXiv (research papers) Neha Nagaraja SI

When Prompts Control Robots: Prompt Injection Attacks in Multi-Agent Robotic Systems

Large language models are increasingly integrated into autonomous robotic systems for task planning and control, but this integration exposes them to prompt injection attacks that can lead to unsafe decisions and physical harm. Multi-agent settings increase the risks through cross-agent contamination and broader attack surfaces. In this paper, we evaluate prompt injection attacks against an LLM-based multi-agent robotic system, considering both direct injections into task…

Read at arXiv (research papers)

arXiv (research papers) Jianshuo Dong SI

Your Agentic LLMs Secretly Encode Indirect Prompt-Injection Exposure in Hidden States

Agentic LLMs are vulnerable to indirect prompt injection (IPI) attacks, e.g., malicious side-tasks hidden in external tool results. While many efforts have sought to address this threat, little is known about the internals of agentic LLMs when they are exposed to IPI attacks. For simplicity, we refer to this condition as IPI exposure. In this paper, we study IPI exposure from three perspectives. (1) Probing: Across eight models, including the 753B-parameter GLM-5.2 and the…

Read at arXiv (research papers)

arXiv (research papers) Minghui Pan SI

Tool Specifications Matter: Uncovering and Mitigating Safety Risks in AI Agents

AI agents extend large language models (LLMs) with external tools, enabling them to perform complex tasks and translate model outputs into consequential real-world actions. Yet LLMs often become substantially less safe when deployed as agents, and the source of this degradation remains poorly understood. In this paper, we identify schema-formatted tool specifications as a primary source of agent safety degradation and show, through white-box representation analysis, that…

Read at arXiv (research papers)

arXiv (research papers) Fazhong Liu SI

Security of World-Model-Based Embodied AI: A Lifecycle of Threats, Defenses, and Evaluation

World models give embodied AI a predictive core: they compress observations into states, simulate action-conditioned futures, and enable planning beyond reactive control. This predictive layer, however, opens a new security boundary-compromise can propagate from data, sensors, prompts, or feedback into physical action. Rather than treating world models as an isolated component, this survey traces threats across their entire lifecycle-from data construction and representation…

Read at arXiv (research papers)

arXiv (research papers) Mingxiao Liu Hangzhou Dianzi University SI

Piggybacking on Perception: Stealthy Concurrent Audio Prompt Injections against Multimodal LLM Agents

Large Language Model (LLM)-driven multimodal agents are increasingly deployed to execute autonomous tasks via continuous audio interaction. While this paradigm enhances interaction naturalness, it introduces a critical yet under-explored attack surface, as audio inputs inevitably contain environmental noise beyond user control. In this paper, we investigate concurrent audio prompt injection attacks targeting multimodal agents. Distinct from traditional acoustic attacks on…

Read at arXiv (research papers)

arXiv (research papers) Eric Wallace SI

GPT-Red: Automated Red Teaming via Self-Play at Scale

We introduce \textbf{GPT-Red}, an automated red-teaming agent that is trained to discover novel prompt injection attacks against frontier LLMs. The goal of this model is to evaluate and improve the robustness of our production systems. To this end, we use it to adversarially train GPT-5.6, our most robust model to prompt injections to date. To create GPT-Red, we design a scalable self-play algorithm where the model is tasked with attacking a diverse population of…

Read at arXiv (research papers)

arXiv (research papers) Haowen Dai SI

SafeFlow: Semantic Information-Flow Control for Blocking Malicious Propagation in Multi-Agent Systems

Multi-agent systems improve capability through task decomposition and role specialization, but these same mechanisms introduce an important safety blind spot: a harmful objective can be fragmented into locally plausible subtasks, allowing malicious intent to evade detection by any single agent. This is a growing social-impact challenge: systems handling sensitive information or consequential tools can turn routine delegation into unauthorized disclosure or unsafe action. We…

Read at arXiv (research papers)

arXiv (research papers) Arseny Kravchenko SI

APPA: Recoverable Information-Flow Control for Real-World LLM Agents

LLM agents deployed in practical workflows routinely mix private context, untrusted tool and web outputs, and external side effects. While information-flow control (IFC) provides structural defenses against prompt injection, data exfiltration, and confused-deputy attacks, conventional IFC relies on monotone taint tracking that either over-blocks benign operations or permanently strands downstream execution once an agent ingests unvetted data. We present APPA (Agentic…

Read at arXiv (research papers)

arXiv (research papers) Max Landauer SI

Just Testing, Move Along: Evasion of LLM-based System Log Interpretation by Prompt Injection

Large Language Models (LLMs) are increasingly integrated into Security Operations Center (SOC) workflows, where they support analysts in tasks such as the interpretation of system logs. However, the ability of LLMs to directly process untrusted textual input also introduces new attack surfaces. In particular, attackers can inject contextual information or explicit instructions into log entries in order to influence how malicious activity is interpreted by the model. Despite…

Read at arXiv (research papers)

arXiv (research papers) Mohan Manivannan SI

Agentic Cloud Decoys: A Deception-Driven Framework for Autonomous Intrusion Investigation

Cloud telemetry arrives at a scale that, paradoxically, makes intrusion understanding harder rather than easier. Attackers operate through legitimate identity, federated session tokens, and cloud native APIs indistinguishable from routine administration, and analysts spend an incident reconstructing context the logs already contain. We present Cloud Decoy AI Agent, a framework pairing a high fidelity cloud decoy with an autonomous language model agent that compresses the…

Read at arXiv (research papers)

arXiv (research papers) Zhaoxi Zhang SI

Are You Still the Agent I Authorized? Earned Authority under a Fixed Ceiling for Evolving Agents

Long-lived AI agents increasingly evolve after deployment by retaining experience, acquiring skills and tools, revising workflows, delegating work, and moving across task phases. This improves adaptation but creates a distinct authorization problem. Tool-enabled agents can turn model errors and prompt injections into consequential external actions; when evolution occurs under a live grant, the subject exercising that authority or the context in which it acts may no longer…

Read at arXiv (research papers)

arXiv (research papers) Nikolaos Kekatos SI

Mission-Level Runtime Assurance for LLM-Assisted ISR Swarms over a Verification-Aware Fabric

Swarms of LLM-assisted autonomous robots are increasingly proposed for cooperative intelligence, surveillance, and reconnaissance (ISR) in contested environments. A growing class of their assurance failures arises not within any single platform but across the swarm: individually-compliant actions compose into a mission-level violation: a prohibited objective split across platforms to evade per-platform limits, or a collective budget quietly exceeded. Per-platform guardrails…

Read at arXiv (research papers)

arXiv (research papers) Jay Barach SI

ResumeShield: Channel Separation and an Open Benchmark for Indirect Prompt Injection in AI Resume Screening

An AI resume screener reads a document supplied by the person it is evaluating, inverting the usual trust relationship between an assessor and the material it assesses. Candidates exploit this by concealing instructions inside a resume using white text, zero font size, hidden elements, markup comments, document metadata, or zero width characters. A human reviewer sees nothing, while a naive extraction pipeline places the concealed text into the model prompt, where it is read…

Read at arXiv (research papers)

arXiv (research papers) Vincent Siu SI

Agent Security Needs Redefinition through a Holistic Framework

Agent security is widely treated as a question about action content. Defenses ask whether an instruction looks malicious. Benchmarks ask whether an agent performs a harmful sounding action. \textbf{We argue that agent security is fundamentally a contextual problem, and that the current content based framing systematically misdefines it.} A command to ``delete user data'' might be a routine administrative request or a prompt injection attacking production systems, and the…

Read at arXiv (research papers)

arXiv (research papers) Tejaswi V. Panchagnula SI

Inference-Time Mitigation of Adversarial Political Bias in Large Language Models

As Large Language Models (LLMs) become the mainstay for information retrieval and summarization tasks, ensuring that they are always non-partisan and invulnerable to political bias is a critical step towards safer and more trustworthy Artificial Intelligence (AI). Current model alignment paradigms, such as reinforcement learning from human feedback (RLHF), make LLMs follow overarching safety instructions. However, this instruction tuning can be exploited via adversarial…

Read at arXiv (research papers)

arXiv (research papers) Yedidel Louck SI

Protocol-Level Attacks on Agentic Commerce Platforms: A Cross-Platform Taxonomy, AIP-Bench, and Unified Defense

Agentic commerce platforms let AI agents autonomously discover services, move payments, and wield user credentials on their users' behalf, and they already handle real money. Their security has so far been studied almost entirely at the level of the AI model, through prompt injection and misalignment. We show that the more consequential risks lie one layer down, in the protocol between agents and commerce services. There, vulnerabilities are structural : exploitation is…

Read at arXiv (research papers)

arXiv (research papers) Zhetong Zhang SI

V-DEAL: Diagnosing Video Safety De-Calibration as an Understanding-Refusal Coupling Failure

As Video Large Language Models are increasingly deployed in real-world applications, ensuring their safety alignment has become critical. Counterintuitively, we find that harmful videos paired with benign queries achieve higher attack success rates than the same videos paired with explicitly harmful queries. To understand the underlying mechanism of this vulnerability, we present V-DEAL, a three-level diagnostic framework that jointly analyzes this failure across model…

Read at arXiv (research papers)

arXiv (research papers) Yanyu Chen SI

Reinforcement Learning for Large Language Model Selective Evidence Adoption from Contaminated Retrieval Results

Retrieval-augmented large language models frequently face contexts that interleave useful evidence with misleading statements or instruction-like content. Blanket refusal discards valid evidence, whereas uncritical adoption yields incorrect or unsafe answers. The ability to selectively adopt relevant information while rejecting deceptive or harmful content is therefore critical for reliable deployment in real-world retrieval settings. We introduce SelectBench, a controlled…

Read at arXiv (research papers)

arXiv (research papers) Or Zion Eliav SI

Know Your Agent: Reconnaissance-Driven Pentesting of AI Agents

Traditional pentesting uses reconnaissance at each step to uncover unseen weaknesses, build stronger attacks, and advance the objective; we argue that AI agents require the same treatment. We formalize agent reconnaissance by modeling the process and identifying the knowledge assets it seeks to extract: what they are, how they are used, and which agent weaknesses they exploit to give adversaries leverage in indirect prompt injection attacks. We instantiate these insights in…

Read at arXiv (research papers)

arXiv (research papers) Zhanhao Hu SI

Twin Agent: Context Residual Compression for Privilege Separated Agents

Large language model (LLM) agents are vulnerable to security risks, such as prompt injection attacks from untrusted context that manipulate downstream reasoning and tool use. Existing secure-by-design approaches mitigate this risk by separating untrusted observations from privileged execution and careful control of information flow, but often degrade utility and require extensive task-specific engineering. We thus propose Twin Agent, a general privilege separation design…

Read at arXiv (research papers)

Older stories