From Announcement to Adoption

Abstract

Organizations are investing rapidly in generative artificial intelligence (AI). Licenses are purchased. Policies are published. Courses fill the learning platform. Yet many employees still hesitate at the point that matters: using AI in real work, with sound judgment and a clear sense of responsibility. This evidence-informed practitioner perspective brings Albert Bandura’s work on self-efficacy and modeling to a practical workplace question: How can leaders help employees believe they can use AI capably and responsibly? The article argues that adoption does not begin with an announcement. It grows when leaders make their own learning visible, useful, and safe for others to try. Modeling is the observable leadership practice, self-efficacy is the primary psychological mechanism, and responsible AI adoption is the intended outcome. The MODEL framework translates this argument into five behaviors: Make learning visible, Open the black box, Develop capability together, Experiment on work that matters, and Legitimize reflection and restraint. Each behavior includes a workplace illustration and a responsible-use boundary. The illustrations include naming an AI tool to lower the emotional threshold for early experimentation, protecting time for leaders and teams to learn together, and using AI-supported action learning to examine a real problem. They are examples, not empirical findings. The article offers leaders and talent-development practitioners a theory-anchored guide for helping people navigate the human transition to AI-enabled work.