Search papers, labs, and topics across Lattice.
This paper explores LLM-mediated web attacks, where user input manipulated by large language models can exploit traditional web application vulnerabilities. By systematizing various attack types, such as LLM2SQLi and LLM2XSS, the authors reveal that the LLM serves as a mediation layer, amplifying the risk of backend actions influenced by attacker-controlled input. The experimental case study, TicketOracle, demonstrates significant model-specific variations in susceptibility to these attacks, highlighting the interplay between application architecture and LLM behavior.
LLMs can inadvertently become conduits for web exploitation, transforming benign user input into dangerous backend actions.
Large Language Models are increasingly integrated into web applications through chatbots, tool-calling pipelines, and agentic workflows. In these systems, user input may influence not only generated text, but also backend actions such as database queries, HTTP requests, file operations, template rendering, or API calls. This paper introduces LLM-mediated web attacks, a class of attacks in which attacker-controlled input is transformed by an LLM-integrated application and then reaches traditional web-application sinks. We systematize this attack surface through representative LLM2X variants, including LLM2SQLi, LLM2XSS, LLM2SSTI, LLM2CommandInjection, LLM2IDOR, LLM2CSRF, LLM2XXE, and LLM2SSRF. Our analysis shows that the LLM usually does not create the underlying vulnerability itself; rather, it acts as a mediation layer, and in some tool-enabled settings as a confused deputy, carrying attacker influence into components that trust model-generated or model-influenced content. As an experimental case study, we implement TicketOracle, a Flask-based LLM-integrated web application for evaluating LLM2SSRF across five attack scenarios and seven LLMs. Our results show substantial variation in susceptibility across models, suggesting that exploitation depends both on insecure application architecture and model-specific behavior. We conclude with mitigation strategies across the prompt, model, application, and network layers.
