---
id: 4658
title: "Offensive Security with Large Language Models (2)"
source_domain: theori.io
source_url: "https://theori.io/blog/offensive-security-with-large-language-models-2-48547"
published_at: "2024-09-30 09:00"
tags: ["ai-security", "prompt-injection", "offensive-security", "sast", "llm", "static-analysis", "codeql"]
phying_url: "https://phying.com/item/4658"
lang: en
---

# Offensive Security with Large Language Models (2)

## Summary

Part two of Xint’s LLM offensive-security series: how LLMs assist static analysis where CodeQL falls short on third-party taint specs, plus injection risks in LLM-integrated web apps.

## Intelligence Metadata

- **Sources:** theori.io
- **Published:** 2024-09-30 09:00
- **Original:** https://theori.io/blog/offensive-security-with-large-language-models-2-48547
- **Phying News:** https://phying.com/item/4658