Global Journal of Engineering and Technology Research (GJETR) · Research article · Open access

Accelerating National Defense with Large Language Models: A Conceptual Framework for the Real-Time Processing of Shared Cyber Threat Indicators

Chukwunenye Amadi, Abolaji Adebayo, Ayokunle Olamide Ijagbemi

5 September 2026 · Vol. 2, Issue 9, pp. 406-427

DOI: 10.65150/EP-gjetr/V2E9/2026-04

Download full text (PDF)Cite this article

Abstract

National defense increasingly depends on the capacity to detect, interpret, and act upon vast streams of shared cyber threat intelligence in real time. Yet defense and security organizations are overwhelmed by unstructured, redundant, and fragmented threat data drawn from heterogeneous global sources, a condition that undermines timely and accurate analysis. This paper develops a conceptual framework for applying Large Language Models (LLMs) to the real-time processing of shared threat indicators within national defense intelligence systems. Grounded in Information Processing Theory, Sociotechnical Systems Theory, and Signal Detection Theory, the framework positions LLMs as advanced cognitive processors that ingest, contextualize, and prioritize language-based threat data while preserving human judgment and accountability. This paper reports no new experimental results. It advances a conceptual argument, supported by an illustrative and explicitly non-experimental comparison, that transformer-based semantic processing is better suited in principle than conventional keyword-matching pipelines to the interpretive demands of heterogeneous and redundantly reported threat intelligence. The comparison is expository in purpose and is offered to generate testable hypotheses rather than to establish validated performance claims. The paper contributes an integrated, governance-aware architecture for intelligent defense systems and articulates the ethical, organizational, and policy conditions under which such systems can be responsibly deployed. The framework yields the testable propositions that LLM-driven semantic processing may reduce analyst cognitive burden, shorten response cycles, and strengthen national cybersecurity posture. Each of these propositions remains to be established through instrumented empirical evaluation, which the paper identifies as the immediate priority for future work.

Keywords: Large Language Models, national defense, cyber threat intelligence, information overload, situational awareness, responsible AI.

Chukwunenye Amadi, Abolaji Adebayo, & Ayokunle Olamide Ijagbemi (2026). Accelerating National Defense with Large Language Models: A Conceptual Framework for the Real-Time Processing of Shared Cyber Threat Indicators. Global Journal of Engineering and Technology Research, 2(9), 406-427. https://doi.org/10.65150/EP-gjetr/V2E9/2026-04

@article{Chukwunenye2026,
  title   = {Accelerating National Defense with Large Language Models: A Conceptual Framework for the Real-Time Processing of Shared Cyber Threat Indicators},
  author  = {Chukwunenye Amadi and Abolaji Adebayo and Ayokunle Olamide Ijagbemi},
  journal = {Global Journal of Engineering and Technology Research},
  year    = {2026},
  volume  = {2},
  number  = {9},
  pages   = {406-427},
  doi     = {10.65150/EP-gjetr/V2E9/2026-04},
  url     = {https://doi.org/10.65150/EP-gjetr/V2E9/2026-04}
}

← All articlesAbout GJETR and how to submit