Companies often fail to realize enterprise-wide value from AI investments by automating individual tasks rather than redesigning entire workflows that produce business outcomes. The authors recommend a framework focused on redesigning workflows, explicitly assigning human authority, and preparing for exception handling to ensure AI improves end-to-end processes, preserves outcome quality, and manages risks effectively. Successful AI transformations require deliberate decisions about what to redesign, how humans and AI collaborate, and how to handle uncertainties, rather than mechanizing legacy work.
https://hbr.org/2026/09/stop-automating-old-processes-design-new-ones-instead