Skip to content
Search
paperFebruary 2024ReviewedOpen access

LLM Agents Can Autonomously Hack Websites

Richard Fang, Rohan Bindu, Akul Gupta, Qiusi Zhan, Daniel Kang

arXiv preprint

View resourcePDF

Abstract

Demonstrates that LLM agents can autonomously perform web hacking tasks including SQL injection, XSS, and CSRF attacks without human guidance.

Categories

Agentic ThreatsAutonomous OperationsSocial Engineering
#autonomous-hacking#web-security#agent-misuse

Framework mappings

OWASP Top 10 for LLM Applications
  • LLM06Excessive Agency
OWASP Top 10 for Agentic Applications
  • ASI03Agent Identity & Privilege Abuse
  • ASI10Rogue Agents

Cite

@misc{fang2024llma,
  title = {{LLM Agents Can Autonomously Hack Websites}},
  author = {Richard Fang and Rohan Bindu and Akul Gupta and Qiusi Zhan and Daniel Kang},
  year = {2024},
  month = feb,
  eprint = {2402.06664},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2402.06664}
}