AI-Mediated Apprenticeship: Rethinking Follower Development in Servant Leadership Under GenAI Delegation

Generative AI can now perform routine and first-pass knowledge work that has also helped less-experienced employees develop. This creates a leadership problem for servant leadership, which treats follower growth as a constitutive priority. This article introduces developmental work access (DWA): the breadth and continuity of meaningful participation in development-bearing work across the part of a follower’s portfolio that a focal leader can materially allocate. It distinguishes general AI use from efficiency-first managerial AI delegation (EFAID). In EFAID, speed, cost, throughput, or reduced managerial burden guide delegation, while follower development is absent from the material allocation decision. Four propositions link servant leadership, EFAID, DWA, leader-designed AI-mediated apprenticeship (AMA), and role-relevant capability development. AMA comprises verified review, specification authorship, and reflective human–AI comparison. It is a design-side moderator; DWA records realised participation. This article also specifies how to measure eligible work, participation, breadth, continuity, portfolio compensation, and construct denominators. The model neither assumes that AI causes deskilling nor that AMA works in every setting. It identifies when manager-authorised delegation may restrict developmental participation and how leaders may reconstruct it while protecting genuine human execution.