arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2408.03515cs.ROcs.AI
收藏

针对集成 LLM 的移动机器人系统的提示注入攻击研究

A Study on Prompt Injection Attack Against LLM-Integrated Mobile Robotic Systems

The University of Western Australia(西澳大利亚大学)

Wenxiao Zhang, Xiangrui Kong, Conan Dewitt, Thomas Braunl, Jin B. Hong

AI总结:

本文研究集成 LLM 的移动机器人面临的多模态提示注入风险,评估安全提示与鲁棒防御机制,使攻击检测和系统性能整体提升约 30.8%。

摘要:

将 GPT-4o 等大型语言模型(LLM)集成到机器人系统中,代表了具身人工智能的一项重大进展。这些模型能够处理多模态提示,从而生成更具情境感知能力的响应。然而,这种集成并非没有挑战。主要担忧之一是在机器人导航任务中使用 LLM 所伴随的潜在安全风险。这些任务需要精确、可靠的响应,以确保安全有效的运行。多模态提示虽然增强了机器人的理解能力,但也带来了可能被恶意利用的复杂性。例如,旨在误导模型的对抗性输入可能导致错误或危险的导航决策。本研究调查了在集成 LLM 的系统中,提示注入对移动机器人性能的影响,并探索用于缓解这些风险的安全提示策略。研究结果表明,通过实施鲁棒防御机制,攻击检测和系统性能两方面均实现了约 30.8% 的显著总体提升,凸显了其在增强面向任务的安全性与可靠性方面的关键作用。

英文摘要:

The integration of Large Language Models (LLMs) like GPT-4o into robotic systems represents a significant advancement in embodied artificial intelligence. These models can process multi-modal prompts, enabling them to generate more context-aware responses. However, this integration is not without challenges. One of the primary concerns is the potential security risks associated with using LLMs in robotic navigation tasks. These tasks require precise and reliable responses to ensure safe and effective operation. Multi-modal prompts, while enhancing the robot's understanding, also introduce complexities that can be exploited maliciously. For instance, adversarial inputs designed to mislead the model can lead to incorrect or dangerous navigational decisions. This study investigates the impact of prompt injections on mobile robot performance in LLM-integrated systems and explores secure prompt strategies to mitigate these risks. Our findings demonstrate a substantial overall improvement of approximately 30.8% in both attack detection and system performance with the implementation of robust defence mechanisms, highlighting their critical role in enhancing security and reliability in mission-oriented tasks.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑