· Blog
Should every strategic decision come with a record of what the AI did?
Aviation has the black box. Medicine has the patient file. What would a decision record for AI-assisted strategy look like?
When an aircraft lands, its flight data are recorded. When a doctor prescribes, the decision is documented. When an executive approves a strategy shaped by AI, there is usually no trace of what the system contributed.
I am working on what such a record might contain, and on a harder question: whether documenting AI use protects the quality of thinking, or merely makes it auditable. To explore it, I combine design science with structured rounds of expert consultation, asking practitioners and scholars what they would need to see, and what they would rather not have to write down.
The answer will matter for regulators, for boards and for anyone who will one day be asked: why did you decide this?