Out of the Cage: How Stochastic Parrots Win in Cyber Security Environments

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F24%3A00372898" target="_blank" >RIV/68407700:21230/24:00372898 - isvavai.cz</a>

Result on the web

<a href="https://doi.org/10.5220/0012391800003636" target="_blank" >https://doi.org/10.5220/0012391800003636</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.5220/0012391800003636" target="_blank" >10.5220/0012391800003636</a>

Result language

angličtina

Original language name

Out of the Cage: How Stochastic Parrots Win in Cyber Security Environments

Original language description

Large Language Models (LLMs) have gained widespread popularity across diverse domains involving text generation, summarization, and various natural language processing tasks. Despite their inherent limitations, LLM-based designs have shown promising capabilities in planning and navigating open-world scenarios. This paper introduces a novel application of pre-trained LLMs as agents within cybersecurity network environments, focusing on their utility for sequential decision-making processes. We present an approach wherein pre-trained LLMs are leveraged as attacking agents in two reinforcement learning environments. Our proposed agents demonstrate similar or better performance against state-of-the-art agents trained for thousands of episodes in most scenarios and configurations. In addition, the best LLM agents perform similarly to human testers of the environment without any additional training process. This design highlights the potential of LLMs to address complex decision-making tasks within cybersecurity efficiently. Furthermore, we introduce a new network security environment named NetSecGame. The environment is designed to support complex multi-agent scenarios within the network security domain eventually. The proposed environment mimics real network attacks and is designed to be highly modular and adaptable for various scenarios.

Czech name

—

Czech description

—

Type

D - Article in proceedings

CEP classification

—

OECD FORD branch

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Project

<a href="/en/project/VJ02010020" target="_blank" >VJ02010020: AI-Dojo: Multiagent Testbed for Research and Testing of AI-driven Cybersecurity Technologies</a><br>

Continuities

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Publication year

2024

Confidentiality

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Article name in the collection

Proceedings of the 16th International Conference on Agents and Artificial Intelligence (Volume 3)

ISBN

978-989-758-680-4

ISSN

2184-3589

e-ISSN

2184-433X

Number of pages

8

Pages from-to

774-781

Publisher name

Science and Technology Publications, Lda

Place of publication

Setúbal

Event location

Rome

Event date

Feb 24, 2024

Type of event by nationality

WRD - Celosvětová akce

UT code for WoS article

—