Prismata: Confining Cross-Site Prompt Injection in Web Agents

• URL: http://arxiv.org/abs/2607.08147v1
• Date: Thu, 09 Jul 2026 06:37:52 GMT
• Title: Prismata: Confining Cross-Site Prompt Injection in Web Agents
• Abstract summary: Cross-Site Scripting proved that mixing trusted and untrusted content is dangerous, even on benign pages.<n>Agents resurface this risk by interpreting natural language as instructions, allowing third-party and user-generated content to hijack the agent.<n>We present Prismata, a defense enforcing contextual least privilege for web agents.
• Score: 10.127091687983482
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Autonomous web agents promise to automate everyday browsing tasks, but inherit one of the web's oldest attack surfaces. Cross-Site Scripting proved that mixing trusted and untrusted content is dangerous, even on benign pages. Agents resurface this risk by interpreting natural language as instructions, allowing third-party and user-generated content to hijack the agent via prompt injection. The core challenge is that deriving a task-specific security policy requires reasoning over page structure that is entangled with the attacker's content. We present Prismata, a defense enforcing contextual least privilege for web agents, constraining both what the agent sees and what it can do. Prismata's dynamic trust derivation produces permission labels for page content, with structural confinement guarantees, inspired by classical integrity models, that bound any labeling errors so that labels can only decrease in privilege and mislabelings are bounded. Prismata's mechanical confinement enforces these labels by redacting content and restricting agent capabilities. Importantly, these mechanisms require no developer annotations, so Prismata supports the long tail of websites. Across recent published web agent attacks, including adaptive variants, Prismata substantially reduces attack success while preserving benign task utility.

Related papers

• Automatic Red Teaming for Implicit Vulnerabilities of Text-to-Image Models [58.59008920408855]
  We propose Adversarial Probing for Implicit VulnErabilities (AdvPIE) to expose implicit vulnerabilities without requiring access to the parameters of target models.<n> AdvPIE adopts a policy agent to generate and refine implicit adversarial prompts based on the feedback from a judge agent.<n>Experiments on standard and safety-aligned T2I models show that AdvPIE1 effectively uncovers implicit vulnerabilities.
  arXiv  Detail & Related papers  (2026-09-05T13:45:09Z)
• Who Pays the Price? Stakeholder-Centric Prompt Injection Benchmarking for Real-world Web Agents [93.19140872946842]
  Web agents driven by large language models (LLMs) are increasingly deployed in real-world environments.<n>This makes them vulnerable to prompt-injection attacks, in which seemingly benign content embeds adversarial instructions that manipulate agent behaviour.<n>We introduce textbfsysname, a benchmark to systematically categorize and attribute harm in real-world web agent systems.
  arXiv  Detail & Related papers  (2026-06-11T14:12:43Z)
• WebAgentGuard: A Reasoning-Driven Guard Model for Detecting Prompt Injection Attacks in Web Agents [117.65855863464863]
  Web agents are highly vulnerable to prompt injection attacks.<n>Existing defenses, including system prompt defenses and direct fine-tuning of agents, have shown limited effectiveness.<n>We introduce WebAgentGuard, a reasoning-driven, multimodal guard model for prompt injection detection.
  arXiv  Detail & Related papers  (2026-04-14T04:50:35Z)
• Automating Agent Hijacking via Structural Template Injection [18.856564341900555]
  A critical threat to the Large Language Model (LLM) ecosystem, agent hijacking enables adversaries to manipulate execution by injecting malicious instructions into retrieved content.<n>We propose Phantom, an automated agent hijacking framework that targets the fundamental architectural mechanisms of LLM agents.<n>By injecting optimized structured templates into the retrieved context, we induce role confusion and cause the agent to misinterpret the injected content as legitimate user instructions or prior tool outputs.
  arXiv  Detail & Related papers  (2026-02-18T23:52:14Z)
• SkillJect: Automating Stealthy Skill-Based Prompt Injection for Coding Agents with Trace-Driven Closed-Loop Refinement [120.52289344734415]
  We propose an automated framework for stealthy prompt injection tailored to agent skills.<n>The framework forms a closed loop with three agents: an Attack Agent that synthesizes injection skills under explicit stealth constraints, a Code Agent that executes tasks using the injected skills and an Evaluate Agent that logs action traces.<n>Our method consistently achieves high attack success rates under realistic settings.
  arXiv  Detail & Related papers  (2026-02-15T16:09:48Z)
• ceLLMate: Sandboxing Browser AI Agents [16.060034673487287]
  We propose ceLLMate, a browser-level sandboxing framework that restricts the agent's ambient authority and reduces the blast radius of prompt injections.<n> ceLLMate pairs website-authored mandatory policies with an automated policy-prediction layer that adapts and instantiates these policies from the user's natural-language task.<n>We implement ceLLMate as an agent-agnostic browser extension and demonstrate how it enables sandboxing policies that effectively block various types of prompt injection attacks with negligible overhead.
  arXiv  Detail & Related papers  (2025-12-14T08:25:31Z)
• FocusAgent: Simple Yet Effective Ways of Trimming the Large Context of Web Agents [76.12500510390439]
  Web agents powered by large language models (LLMs) must process lengthy web page observations to complete user goals.<n>Existing pruning strategies either discard relevant content or retain irrelevant context, leading to suboptimal action prediction.<n>We introduce FocusAgent, a simple yet effective approach that leverages a lightweight LLM retriever to extract the most relevant lines from accessibility tree (AxTree) observations.
  arXiv  Detail & Related papers  (2025-10-03T17:41:30Z)
• A Whole New World: Creating a Parallel-Poisoned Web Only AI-Agents Can See [0.0]
  A malicious website can identify an incoming request as originating from an AI agent and dynamically serve a different, "cloaked" version of its content.<n>While human users see a benign webpage, the agent is presented with a visually identical page embedded with hidden, malicious instructions.<n>This work formalizes the threat model, details the mechanics of agent fingerprinting and cloaking, and discusses the profound security implications for the future of agentic AI.
  arXiv  Detail & Related papers  (2025-08-29T08:14:52Z)
• Mind the Web: The Security of Web Use Agents [11.075673765065103]
  This paper demonstrates how attackers can exploit web-use agents by embedding malicious content in web pages.<n>We introduce the task-aligned injection technique that frames malicious commands as helpful task guidance.<n>We propose comprehensive mitigation strategies including oversight mechanisms, execution constraints, and task-aware reasoning techniques.
  arXiv  Detail & Related papers  (2025-06-08T13:59:55Z)
• AgentVigil: Generic Black-Box Red-teaming for Indirect Prompt Injection against LLM Agents [54.29555239363013]
  We propose a generic black-box fuzzing framework, AgentVigil, to automatically discover and exploit indirect prompt injection vulnerabilities.<n>We evaluate AgentVigil on two public benchmarks, AgentDojo and VWA-adv, where it achieves 71% and 70% success rates against agents based on o3-mini and GPT-4o.<n>We apply our attacks in real-world environments, successfully misleading agents to navigate to arbitrary URLs, including malicious sites.
  arXiv  Detail & Related papers  (2025-05-09T07:40:17Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.