Athanasios Trantas

Large Language Models (LLMs) are rapidly entering defense and security applications, from tactical decision support and wargaming to autonomous uncrewed systems. It is unclear how these models are applied across the defense and security domain, which technical approaches predominate, and how mature the reported work is. Existing surveys cover defense AI broadly or LLM capabilities generally, but no PRISMA-based systematic review has mapped this domain's empirical record against a structured classification scheme. We systematically review the peer-reviewed empirical literature across four databases and characterize every included study with a 20-characteristic codebook spanning military application context, technical implementation, and responsible AI practices. The 56 included studies (2023 to March 2026) form a gradient from descriptive analytics to agentic and autonomous systems, with systems holding direct control authority remaining a minority (9 of 56). The technical profile is shallow: 91% of studies are proofs of concept, prompt-engineered commercial models dominate, and responsible AI safeguards are adopted only sparsely. The codebook gives defense researchers and policymakers a reusable instrument for classifying LLM applications in defense and security, and the identified gaps mark where research and governance attention is needed next.

Contract automation is a challenging topic within Artificial Intelligence and LegalTech. From digitised contracts via smart contracts, we are heading towards Intelligent Contracts (iContracts). We will address the main challenge of iContracts: the handling of communications and risk data in contract automation. Our Research Question reads: to what extent is it possible to develop an ontology that automates contracts with communications and risk data? The article designs and conceptualises an iContract ontology. Our findings validate the conceptual expressiveness of our ontology. A brief discussion highlights the value of the ontology design and its application domains. From the results, we may conclude that the current method is innovative and that further research is necessary for handling more complex use cases.

Course Of Action (COA) generation is an important step in the military decision making process. In land operations it is a manual, creative process throughout which a commander and his staff take many factors into account, such as mission, terrain and weather, doctrine, threat capabilities and own capabilities. Disadvantages of this manual process are the inability of taking into account too many factors, the number of staff involved and the difficulty in evaluating COAs. New (AI-based) technologies can overcome these disadvantages and when implemented could serve as Tactical Decision Aids. This paper presents the results of research in which a step towards an automated tactical decision support tool for land operations was made. For a single SEIZE task at company level, based on doctrinal templates, COAs are selected and optimized using Genetic Algorithms based on simulated effects. An initial step in elucidating the generated COAs is undertaken by linking individual outcomes of simulation runs during the (GA) optimization process to tactical features of the terrain. ^{1 1} This paper was originally presented at the NATO Science and Technology Organization Symposium (ICMCIS) organized by the Information Systems Technology (IST) Panel, IST-205-RSY - the ICMCIS, held in Koblenz, Germany, 23–24 April 2024.