The Black Box, Unlocked: Predictability and Understandability in Military AI
The DOI matches its Crossref record.
Abstract
Predictability and understandability are widely held to be vital characteristics of artificially intelligent systems. Put simply: AI should do what we expect it to do, and it must do so for intelligible reasons. This consideration stands at the heart of the ongoing discussion about lethal autonomous weapon systems and other forms of military AI. But what does it mean for an intelligent system to be "predictable" and "understandable" (or, conversely, unpredictable and unintelligible)? What is the role of predictability and understandability in the development, use, and assessment of military AI? What is the appropriate level of predictability and understandability for AI weapons in any given instance of use? And how can these thresholds be assured? This study provides a clear, comprehensive introduction to these questions, and proposes a range of avenues for action by which they may be addressed.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Recent HCI studies on LLM-based tool design have mainly focused on prompt engineering [6,16,46,55,94,98] and exploring the generative capabilities of LLMs [50,58,63,67,97].”
From ABScribe: Rapid Exploration & Organization of Multiple Writing Variations in Human-AI Co-Writing Tasks using Large Language Models (2024) · cites this paper as [63]
Citations
Cites
Still being checked.
Cited by
Evaluating the quality of machine learning explanations: a survey on methods and metrics
Zhou, J., Gandomi, A. H., Chen, F., & Holzinger, A. (2021) · Electronics · doi:10.3390/electronics10050593 · Cited by 606
“It is argued that two factors determine understandability of an ML system: the features of the ML system and the human’s capacity for understanding [62].”
BackgroundReza, M., Laundry, N. M., Musabirov, I., Dushniku, P., Yu, Z. Y. “., Mittal, K., Grossman, T., Liut, M., Kuzminykh, A., & Williams, J. J. (2024) · Proceedings of the CHI Conference on Human Factors in Computing Systems · doi:10.1145/3613904.3641899 · Cited by 52
“Unlike traditional menu buttons which have predictable outcomes, AI Modifiers can yield unexpected results due to the inherent unpredictability of AI models [40, 64, 95, 96].”
Hidden Risks: Artificial Intelligence and Hermeneutic Harm
Rebera, A. P., Lauwaert, L., & Oimann, A.-K. (2025) · Minds and Machines · doi:10.1007/s11023-025-09733-0 · Cited by 6
“…as (1), where decisions are relatively momentous, XAI techniques enable the development of systems based on white-or grey-box models or which, if based on black-box models, incorporate additional tools to facili - tate interpretation and/or explanation (Holland Michel, 2020 ; Wachter et al., 2018).”
Metadata
- Authors
- United Nations Institute for Disarmament Research, & Holland Michel, A.
- Published
- 2020-09-22
- OpenAlex
- W4207068985
- Cited by
- 64
- References
- 0
- Citations by year
- 2026: 12 · 2025: 9 · 2024: 24 · 2023: 10 · 2022: 5 · 2021: 4
- Type
- report
- Language
- en
- Open access
- closed
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Physical Sciences › Engineering › Control and Systems Engineering › Military Strategy and Technology
- Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)
- Keywords
- artificial intelligence in warfare, lethal autonomous weapons, predictability, comprehensibility, explainability, black box, AI governance, human control over AI
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: The Black Box, Unlocked: Predictability and Understandability in Military AI. https://app.cochat.ai/papers/10.37559/sectec/20/ai1
- The paper
- United Nations Institute for Disarmament Research, & Holland Michel, A. (2020). The Black Box, Unlocked: Predictability and Understandability in Military AI. https://doi.org/10.37559/sectec/20/ai1
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.