ORCID
Jae Oh: 0000-0002-5842-5189
Shiu-Kai Chin: 0009-0007-5318-923X
Document Type
Article
Date
Summer 6-15-2026
Keywords
trustworthy AI, mission assurance, systems security engineering, STPA-SEC, AI risk management, lifecycle analysis, literature survey, formal verification, NIST SP 800- 160, defense AI
Language
English
Funder(s)
National Security Agency
Funding ID
CON06350
Acknowledgements
This project is partially supported by the National Security Agency.
Disciplines
Electrical and Computer Engineering
Description/Abstract
This report presents a systematic analysis of 54,628 AI research papers published between 2021 and 2025, examining the extent to which current government, academic, and industry efforts address the trustworthiness requirements of mission-critical AI systems. Papers were classified along two dimensions: organizational tier (from AI model to mission/business pro- cess) using NIST SP 800-39, and systems engineering lifecycle phase using ISO/IEC/IEEE 15288. Classification combined automated LLM-assisted screening with human validation, achieving inter-rater reliability measured by Krippendorff’s α and Gwet’s AC1. Only 16.8% of papers examined addressed any aspect of AI trustworthiness, reliability, or security, and virtually all were concentrated at the lowest abstraction level—algorithm and implemen- tation concerns. No papers substantively addressed mission assurance: the process of en- suring that systems using AI continue to perform mission-essential functions under adverse conditions. This gap is critical for defense applications, where AI systems must operate within a chain of command and under defined rules of engagement. To address this gap, we recommend the System-Theoretic and Technical Operational Risk Management (STORM) methodology, specifically its STPA-SEC component for early-lifecycle mission requirements definition and its Certified Security by Design (CSBD) component for formal verification that system designs satisfy those requirements. Together, these close the loop between com- mander’s intent, system design, and mission assurance.
Recommended Citation
Oh, Jae C.; Chin, Shiu-Kai; Katz, Garrett; Young, William E.; and Clark, Matt, "Literature Landscape of Government, Academic, and Industry Efforts in Responsible, Secure, and Trustworthy AI" (2026). Electrical Engineering and Computer Science - All Scholarship. 253.
https://surface.syr.edu/eecs/253
Source
submission
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.
