Too Smart for Comfort? How Neural Inference Depth Shapes Consumer Adoption of Brain-Computer Interfaces

As firms increasingly introduce consumer-facing neurotechnology products and services, such as brain–computer interface (BCI)-enabled applications, understanding how consumers respond to varying levels of personal data inference becomes critical for market success. This study examines how neural inference depth—defined as the extent to which a technology infers psychologically sensitive information about users—shapes consumer adoption intention in retail and digital service contexts. Across three between-subjects experiments (total N = 1,547), results reveal an inverted-U relationship between inference depth and adoption intention, with moderate levels of inference maximizing consumer acceptance. We further identify a dual-path mechanism underlying this effect: whereas the self-insight value exhibits diminishing returns at higher levels of inference, perceived creepiness increases at an accelerating rate, leading to a systematic shift in pathway dominance. Perceived data control mitigates the negative effect of creepiness without enhancing perceived benefits. Importantly, a design intervention in the form of modular disclosure, which staggers the activation of advanced features, reduced perceived creepiness and increased adoption intention under conditions of deep inference without compromising self-insight value, although it operated only partly through the creepiness pathway. Based on experimental evidence from adult online consumers in China responding to hypothetical BCI scenarios, these findings contribute to the retailing and consumer services literature by conceptualizing consumer data inference as a continuous hierarchical construct and offer directional guidance for firms designing privacy-sensitive, technology-enabled service experiences, pending cross-cultural and field-based replication.