Abstract

Organizational IS implementations often encounter user resistance, which is exacerbated for emergent technologies where novelty and ambiguous use cases amplify perceived threats. Training is a common managerial lever, yet typical one-size-fits-all approaches neglect dispositional heterogeneity in resistance-prone traits (e.g., routine seeking, emotional reactivity, short-term focus, cognitive rigidity). We propose that adaptive learning can reduce resistance when training is tailored to these differences from the outset. However, standard performance-driven adaptivity is limited by cold-start, whereby early training remains uniform until sufficient performance data accumulate. Using an AI-enabled digital twin simulation of building operations, this study conducts a randomized controlled lab experiment comparing (1) non-adaptive training, (2) traditional performance-adaptive training, and (3) profile-initialized adaptive training that personalizes from the outset using a pre-training dispositional survey. We test effects on resistance intention and performance, with perceived threats as a mediator and dispositional traits as moderators. This study advances IS resistance research by theorizing training design as a mechanism for reducing perceived threats during emergent technology implementation. Using adaptive learning as a design framework, we address the cold-start problem in early training experiences for AI-enabled digital twin training.

Download

Share

COinS