TRAINING TILIKUM: LARGE LANGUAGE GENERATIVE ARTIFICIAL INTELLIGENCE AND HOMELAND SECURITY PROGRAM APPLICATIONS

Loading...
Thumbnail Image
Authors
Greenshields, James L.
Subjects
artificial intelligence
AI
emergency management
homeland security
large language models
preparedness
planning
exercise
Advisors
Wollman, Lauren
Brown, Shannon A.
Date of Issue
2024-03
Date
Publisher
Monterey, CA; Naval Postgraduate School
Language
Abstract
This forward-thinking study explores the application of large language model (LLM) generative artificial intelligence (AI) into homeland security programs, investigating their potential to enhance adaptability and responsiveness to complex challenges. This research assesses the effectiveness of LLM through a literature review, thematic analysis, and practical planning and exercise development experiments within a fictional training jurisdiction. Graded rubric criteria, informed through standards and best practices, categorized LLM output effectiveness. Results indicate that while LLMs offer support, particularly in exercise design, their application is contingent upon the quality of prompt development, training data, and situational human input. The study advocates for a cautious adoption strategy, highlighting the importance of ethical considerations, continuous evaluation, and decomposing complex tasks to optimize LLM output. Recommendations include leveraging AI for data-intensive tasks, defining a human-AI collaboration to enhance decision-making and creativity, and using a newly developed Deliberative Implementation Framework for LLM AI to navigate the policy complexities of AI program integration. This work elucidates the nuanced interplay between AI and human capabilities within homeland security, providing a perspective on AI’s potential to transform and enhance programs while cautioning against over-reliance on technology at the expense of human judgment.
Type
Thesis
Description
Series/Report No
Department
National Security Affairs (CHDS)
Organization
Identifiers
NPS Report Number
Sponsors
Funder
Format
Citation
Distribution Statement
Distribution Statement A. Approved for public release: Distribution is unlimited.
Rights
This publication is a work of the U.S. Government as defined in Title 17, United States Code, Section 101. Copyright protection is not available for this work in the United States.
Collections