4,239 rated or chosen profiles from 471 respondents, in up to 9 tasks of 1 profile varying on 9 attributes.

About this table

Description	US Army ROTC cadets (Qualtrics, Nov-Dec 2023) rated their trust in 9 hypothetical AI-enabled military technologies used in a large war, varying platform, purpose, precision, autonomy, civilian deaths, friendly forces saved, mission contribution, regulation and other adopters. 471 respondents; one 1-5 trust rating per profile.
Reference	Lushenko, P. (2025). AI, trust, and the war room: Evidence from a conjoint experiment in the US military. Cambridge Forum on AI: Law and Governance.
DOI	10.1017/cfl.2025.10019
Licence	CC0 1.0
Source data	https://doi.org/10.7910/DVN/KDSDXC

Notes

Conjoint table (irw_conjoint). rating: "Do you trust partnering with this AI-enhanced military technology when it is used under these conditions?" 1 (strong distrust) .. 5 (strong trust), raw; single profile per task, 9 tasks, fixed attribute order. Reliability scenario shown before and after the conjoint dropped (attributes not exported). 3 of 474 respondents dropped (attribute levels not saved). Covariates are unlabelled codes; masked IP/location/name columns and the free-text item dropped.

These notes describe the source data, which the IRW reproduces as released. All notes are listed on the Data Notes page.

Design

Topic	technology ai privacy, foreign policy security
A profile represents	product service
Country	US
Language shown	en
Language of stored levels	en
Task order	recorded in the deposit
Profile position	recorded in the deposit

Outcomes:

rating	Do you trust partnering with this AI-enhanced military technology when it is used under these conditions? [rating, 1 (strong distrust) to 5 (strong trust)]

Columns

attr_autonomyattr_civilian_casualtiesattr_friendly_forces_savedattr_mission_contributionattr_other_adoptersattr_platformattr_precisionattr_purposeattr_regulationcov_age_group_codecov_duration_seccov_education_codecov_gender_codecov_ideology_codecov_military_role_codecov_party_codecov_race_codeidprofileratingtask

The layout (one row per respondent, task and profile; choice/rating outcomes; attr_ columns holding each attribute level as displayed) is defined in the IRW conjoint standard. What the outcomes asked is under Design above.

Get the data

Download CSVno account needed Browse on Redivisexplore and query Croissant metadataHugging Face, Kaggle, OpenML

Or load it directly in R or Python (both need a free Redivis account; the first call opens a browser to sign in):

# R
install.packages("remotes")
remotes::install_github("itemresponsewarehouse/Rpkg")
library(irw)
df <- irw_fetch("lushenko_2025_ai_trust_cadets", source = "conj")

# Python
pip install irw

import irw
df = irw.fetch("lushenko_2025_ai_trust_cadets", source="conj")

How to cite

@article{Lushenko_2025, title={AI, trust, and the war room: Evidence from a conjoint experiment in the US military}, volume={1}, ISSN={3033-3733}, url={http://dx.doi.org/10.1017/cfl.2025.10019}, DOI={10.1017/cfl.2025.10019}, journal={Cambridge Forum on AI: Law and Governance}, publisher={Cambridge University Press (CUP)}, author={Lushenko, Paul}, year={2025} }

Respondents	471
Rows (one per respondent, task and profile)	4,239
Tasks per respondent (most)	9
Profiles per task	1
Attributes	9

IRW version	v584
Redivis dataset	irw_conjoint v4.2
Redivis dataset DOI	10.57761/qkye-q530
Manifest pin for this IRW version	v4.2
Metadata source	irw_meta v44.0

lushenko_2025_ai_trust_cadets