Explainable Artificial Intelligence: State of the Art and beyond

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F60162694%3AG43__%2F26%3A00564758" target="_blank" >RIV/60162694:G43__/26:00564758 - isvavai.cz</a>

Výsledek na webu

<a href="https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11061287" target="_blank" >https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11061287</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/ICMT65201.2025.11061287" target="_blank" >10.1109/ICMT65201.2025.11061287</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Explainable Artificial Intelligence: State of the Art and beyond

Popis výsledku v původním jazyce

Explainable Artificial Intelligence (XAI) is critical in military applications where trust, interpretability, and resilience against adversarial threats are paramount. Despite recent advances, existing XAI methodologies often fail to meet military-grade requirements, particularly in real-Time decision-making, cybersecurity, and classified operational environments. This paper conducts a systematic review of state-of-The-Art XAI methods, evaluating their applicability, limitations, and adversarial robustness in military contexts. We analyzed peer-reviewed publications and defense strategy reports to assess the computational trade-offs, security risks, and feasibility of XAI methods in combat decision support, cyber defense, and autonomous warfare. Our findings highlight three key research gaps: (1) the lack of military-specific XAI benchmarks, (2) the need for cryptographically secured explainability mechanisms to counter adversarial manipulation, and (3) the challenge of balancing transparency with operational security (OPSEC). We propose a three-layered XAI framework to address these issues and outline future research priorities for developing secure, adaptive, and real-Time interpretable AI for defense applications.

Název v anglickém jazyce

Explainable Artificial Intelligence: State of the Art and beyond

Popis výsledku anglicky

Explainable Artificial Intelligence (XAI) is critical in military applications where trust, interpretability, and resilience against adversarial threats are paramount. Despite recent advances, existing XAI methodologies often fail to meet military-grade requirements, particularly in real-Time decision-making, cybersecurity, and classified operational environments. This paper conducts a systematic review of state-of-The-Art XAI methods, evaluating their applicability, limitations, and adversarial robustness in military contexts. We analyzed peer-reviewed publications and defense strategy reports to assess the computational trade-offs, security risks, and feasibility of XAI methods in combat decision support, cyber defense, and autonomous warfare. Our findings highlight three key research gaps: (1) the lack of military-specific XAI benchmarks, (2) the need for cryptographically secured explainability mechanisms to counter adversarial manipulation, and (3) the challenge of balancing transparency with operational security (OPSEC). We propose a three-layered XAI framework to address these issues and outline future research priorities for developing secure, adaptive, and real-Time interpretable AI for defense applications.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10200 - Computer and information sciences

Projekt

—

Návaznosti

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Rok uplatnění

2025

Kód důvěrnosti údajů

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

Název statě ve sborníku

2025 10th International Conference on Military Technologies, ICMT 2025 - Proceedings

ISBN

979-8-3315-2338-1

ISSN

2996-4466

e-ISSN

2996-4474

Počet stran výsledku

7

Strana od-do

—

Název nakladatele

Institute of Electrical and Electronics Engineers Inc.

Místo vydání

—

Místo konání akce

Brno, Czech Republic

Datum konání akce

27. 5. 2025

Typ akce podle státní příslušnosti

CST - Celostátní akce

Kód UT WoS článku

001545807300030