Type of Submission
Poster
Keywords
Artificial Intelligence, Cyber Operations, Cyber Kill Chain, Common Vulnerabilities and Exposures, Knowledge Augmented-Generation, Large Language Model, Multi-agent System, Retrieval-Augmented Generation, Tactics Techniques and Procedures
Proposal
In the cybersecurity field, a significant skills gap exists between senior-level operators and novice operators. The goal of this project is to develop an Artificial Intelligence Offensive Cyber Assistant that enables novice cyber operators to perform at the level of an expert. This assistant is called Multi-Agent AI Enhancing Security, Threat Response, and Orchestration (MAESTRO). Through our research and development, we built a program integrating local or cloud-based state-of-the-art Large Language Models (LLMs) that will give accurate responses to assist a cyber operator in developing tactics. We utilize a Multi-Agent System (MAS) and Knowledge-Augmented Generation (KAG), as well as rigorous user testing, to increase confidence in our system. This research addresses ways to improve accuracy and consistency in LLM responses to confidently use artificial intelligence to scale operators in the cyber operation sector.
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.
Copyright
© 2025 Gabriela J. Mallack. All rights reserved.
MAESTRO: Multi-Agent AI Enhancing Security, Threat Response, and Orchestration
In the cybersecurity field, a significant skills gap exists between senior-level operators and novice operators. The goal of this project is to develop an Artificial Intelligence Offensive Cyber Assistant that enables novice cyber operators to perform at the level of an expert. This assistant is called Multi-Agent AI Enhancing Security, Threat Response, and Orchestration (MAESTRO). Through our research and development, we built a program integrating local or cloud-based state-of-the-art Large Language Models (LLMs) that will give accurate responses to assist a cyber operator in developing tactics. We utilize a Multi-Agent System (MAS) and Knowledge-Augmented Generation (KAG), as well as rigorous user testing, to increase confidence in our system. This research addresses ways to improve accuracy and consistency in LLM responses to confidently use artificial intelligence to scale operators in the cyber operation sector.
