Narrative literature reviews can be subjective, laborious, and difficult to replicate. To minimize human inputs and increase transparency, we generated a thematic scheme of leadership topics, and used software tools to code and organize 2115 leadership articles authored by 3190 authors and published in 10 academic journals from 1990 to 2017. We visually present the empirically-grounded taxonomies of 56 popular research topics and the collaboration network of 160 productive leadership scholars. Shifts of popular topics and active scholars across four 7-year intervals are shown, with our reflections on the rises and declines of charismatic and authentic leadership theories. We discuss how artificial intelligence can enhance the scientific value of literature reviews.