Directors do not need to be technologists. They do need eight competencies, each restated for AI and measured against the oversight duty, when AI becomes central to care, products, and operations.

An AI tool that refines a process and one that changes who makes a clinical, coverage, or regulatory decision should not reach the board in the same form. A four-question test helps directors tell them apart.

Insolvency is the last stage of a sequence that starts in the boardroom. Directors who know the four stages, and where AI speeds them, can interrupt the sequence while options are still wide.

Health system boards do not need to understand how AI models work. They need to know which uses serve how the organization is paid, what each kind of model demands after approval, which standards they will be measured against, and who is accountable.

The best-known corporate failures were rarely failures of technology or foresight. They were failures of what the board saw, asked, and recorded, and AI oversight turns on the same question.

AI has not changed what a board is for. It has changed what the board must see, ask, and record, and where the line with the chief executive falls.

Boards rarely decide with complete information. The failure is using one method for four kinds of gaps. This guide shows directors how to name each gap, match their questions to it, and record what they decided, from R&D programs to AI approvals.

Boards rarely miss AI risk for lack of interest. Here are eight places the board pack does not reach, the evidence behind each, and what directors should ask.