Skip to yearly menu bar Skip to main content

Workshop: Instruction Tuning and Instruction Following

Identifying and Mitigating Vulnerabilities in LLM-Integrated Applications

Fengqing Jiang · Zhangchen Xu · Luyao Niu · Boxin Wang · Jinyuan Jia · Bo Li · Radha Poovendran

Keywords: [ large language model ] [ privacy ] [ toxicity ] [ misuse mitigation ] [ bias ] [ safety of LLM-integrated application ] [ disinformation ]


The remarkable instruction following capabilities of large language models (LLMs) allow them to be increasingly deployed as the service backend for LLM-integrated applications such as code completion and AI-powered search. Compared with the traditional usage of LLMs where users directly send queries to an LLM, LLM-integrated applications serve as middleware to refine users’ queries with domain-specific knowledge to better inform LLMs and enhance the responses. Despite numerous opportunities and benefits, blindly following instructions given to LLMs exposes LLM-integrated applications to new attack surfaces. Understanding, minimizing, and eliminating the emerging attack surfaces is a new area of research. In this work, we consider a setup where the user and LLM interact via an LLM-integrated application in the middle. We focus on the communication rounds that begin with user’s queries and end with LLM-integrated application returning responses to the queries, powered by LLMs at the service backend. For this query-response protocol, we identify potential high-risk vulnerabilities that can originate from the malicious application developer or from an outsider threat initiator that is able to control the database access, manipulate and poison data that are high-risk for the user. Successful exploits of the identified vulnerabilities result in the users receiving responses tailored to the intent of a threat initiator (e.g., biased preferences for certain products). We assess such threats against LLM-integrated applications empowered by OpenAI GPT-3.5 and GPT-4. Our empirical results show that the threats can effectively bypass the restrictions and moderation policies of OpenAI, resulting in users receiving responses that contain bias, toxic content, privacy risk, and disinformation. To mitigate those threats, we identify and define four key properties, namely integrity, source identification, attack detectability, and utility preservation, that need to be satisfied by a safe LLM-integrated application. Based on these properties, we develop a lightweight, threat-agnostic defense that mitigates both insider and outsider threats. Our evaluations demonstrate the efficacy of our defense.

Chat is not available.