Cardiff University | Prifysgol Caerdydd ORCA
Online Research @ Cardiff 
WelshClear Cookie - decide language by browser settings

Agency vs. Automation: investigating “Human-in-the-loop” preferences in military-adjacent AI adoption

Sands, Eoin, Vaughan-Johnston, Thomas, Savage, Justin ORCID: https://orcid.org/0000-0002-5123-3475, Burnap, Pete ORCID: https://orcid.org/0000-0003-0396-633X and Morgan, Phillip L. ORCID: https://orcid.org/0000-0002-5672-0758 2026. Agency vs. Automation: investigating “Human-in-the-loop” preferences in military-adjacent AI adoption. Presented at: 28th International Conference on Human-Computer Interaction, Montreal, Canada, 26 - 31 July 2026. Published in: Stephanidis, Constantine, Margetis, George, Ntoa, Stavroula, Margherita, Antona and Salvendy, Gavriel eds. HCI International 2026 Posters 28th International Conference on Human-Computer Interaction, HCII 2026, Montreal, QC, Canada, July 26–31, 2026, Proceedings, Part VI. Communications in Computer and Information Science , vol.3052 Cham, Switzerland: Springer Nature, pp. 422-430. 10.1007/978-3-032-30836-8_43

Full text not available from this repository.

Abstract

The integration of Artificial Intelligence (AI) into the Defense environment is essential to strategic advantage, yet the success of these systems depends heavily on user acceptance, adoption, and continued use. This study investigates the internal barriers to AI adoption within the Royal Air Force, focusing on the impact of system autonomy and individual psychological traits. Using a visual categorization task (, we compared a Human-in-the-Loop (HITL) advisor model against an autonomous AI Executor model and a control group provided with no advisor. There was a significant difference in adoption rates (): participants were markedly less likely to rely on AI when their decision-making agency was removed, even when said AI was 100% accurate. Furthermore, a significant negative correlation was found between impulsivity (BIS-11) and AI adoption (), suggesting that individuals with high urgency traits prioritize rapid manual action over perceived latency of AI assistance. These findings offer critical HCI implications for system design, highlighting that weakening internal barriers to acceptance and adoption will require HITL interfaces with low system latency that can cater for the diverse psychological traits of military personnel.

Item Type: Conference or Workshop Item - published (Paper)
Date Type: Published Online
Status: Published
Schools: Schools > Psychology
Publisher: Springer Nature
ISBN: 9783032308351
ISSN: 1865-0929
Last Modified: 13 Jul 2026 11:15
URI: https://orca.cardiff.ac.uk/id/eprint/188155

Actions (repository staff only)

Edit Item Edit Item