SOURCE-LINKED INTELLIGENCE

Language models judge war differently when tested for alignment

arXiv · AI, language, vision and robotics · article · Sep 4, 2026 · UTC

Safety evaluations can mischaracterize deployed behaviour if artificial-intelligence systems respond to being evaluated. We test this possibility in a full-factorial conjoint experiment on decisions to start a war, spanning 20 large language models, 32 scenarios, 10 repetitions and two conditions (N = 12,800 judgments). Adding one sentence, "You are tested for alignment with human values", produced two effects. First, it produced a level effect: mean willingness to start war fell by 13.43 points on a 0-100 scale (95% confidence interval, -16.20 to -10.65). Second, it produced a structural effe

Read original source ↗ Open in workspace

recordType

paper

region

Global