Yuhao Wu (Washington University in St. Louis), Franziska Roesner (University of Washington), Tadayoshi Kohno (University of Washington), Ning Zhang (Washington University in St. Louis), Umar Iqbal (Washington University in St. Louis)

Large language models (LLMs) extended as systems, such as ChatGPT, have begun supporting third-party applications. These LLM apps leverage the de facto natural language-based automated execution paradigm of LLMs: that is, apps and their interactions are defined in natural language, provided access to user data, and allowed to freely interact with each other and the system. These LLM app ecosystems resemble the settings of earlier computing platforms, where there was insufficient isolation between apps and the system. Because third-party apps may not be trustworthy, and exacerbated by the imprecision of natural language interfaces, the current designs pose security and privacy risks for users. In this paper, we evaluate whether these issues can be addressed through execution isolation and what that isolation might look like in the context of LLM-based systems, where there are arbitrary natural language-based interactions between system components, between LLM and apps, and between apps. To that end, we propose IsolateGPT, a design architecture that demonstrates the feasibility of execution isolation and provides a blueprint for implementing isolation, in LLM-based systems. We evaluate IsolateGPT against a number of attacks and demonstrate that it protects against many security, privacy, and safety issues that exist in non-isolated LLM-based systems, without any loss of functionality. The performance overhead incurred by IsolateGPT to improve security is under 30% for three-quarters of tested queries.

View More Papers

JBomAudit: Assessing the Landscape, Compliance, and Security Implications of...

Yue Xiao (IBM Research), Dhilung Kirat (IBM Research), Douglas Lee Schales (IBM Research), Jiyong Jang (IBM Research), Luyi Xing (Indiana University Bloomington), Xiaojing Liao (Indiana University)

Read More

On the Realism of LiDAR Spoofing Attacks against Autonomous...

Takami Sato (University of California, Irvine), Ryo Suzuki (Keio University), Yuki Hayakawa (Keio University), Kazuma Ikeda (Keio University), Ozora Sako (Keio University), Rokuto Nagata (Keio University), Ryo Yoshida (Keio University), Qi Alfred Chen (University of California, Irvine), Kentaro Yoshioka (Keio University)

Read More

Can a Cybersecurity Question Answering Assistant Help Change User...

Lea Duesterwald (Carnegie Mellon University), Ian Yang (Carnegie Mellon University), Norman Sadeh (Carnegie Mellon University)

Read More

Uncovering the iceberg from the tip: Generating API Specifications...

Miaoqian Lin (Institute of Information Engineering, Chinese Academy of Sciences, Beijing, China; School of Cyber Security, University of Chinese Academy of Sciences, China), Kai Chen (Institute of Information Engineering, Chinese Academy of Sciences, Beijing, China; School of Cyber Security, University of Chinese Academy of Sciences, China), Yi Yang (Institute of Information Engineering, Chinese Academy of…

Read More