Beyond Optimization: Perceptual Integrity as a Foundation for Human–AI Coherence

As artificial intelligence increasingly mediates decision-making across organizational and societal contexts, a critical challenge emerges concerning the preservation of human cognitive coherence under system-driven conditions. While prior research has emphasized trust, fairness, and transparency, it provides limited explanation of whether individuals remain cognitively aligned with decisions generated through algorithmic processes. This paper introduces perceptual integrity as a construct capturing the extent to which individuals maintain coherence, interpretability, and authorship over their decisions in human–AI interaction. Building on a proposed theory of conscious leadership, cognition is conceptualized as an emergent interaction among environment, memory, systems, and the human agent, where decision quality depends on the balance among these forces. An experimental study (N = 602) was conducted to examine the impact of decision structure on perceptual integrity and its relationship with trust. Participants were assigned to either an algorithmic imposition condition or an interpretive autonomy condition. Results indicate that algorithmic imposition reduces perceptual integrity compared to interpretive autonomy (t(600) = 4.21, p < 0.001, d = 0.38). Furthermore, perceptual integrity significantly predicts trust (β = 0.36, p < 0.001) and partially mediates the relationship between decision condition and trust (indirect effect = 0.17, 95% CI [0.09, 0.27]). These findings suggest that trust in AI-assisted decisions is shaped not only by system performance but also by the degree to which individuals remain cognitively engaged in the decision process. By introducing perceptual integrity as a measurable manifestation of cognitive balance, this study contributes to human–AI research and offers a new perspective on leadership as the management of cognitive conditions rather than behavioral outcomes.