AI has moved from the innovation lab to the boardroom. Directors are now expected to exercise informed oversight of AI risk and opportunity. Five questions separate mature boards from the rest.
First: what decisions will this system influence, and who is accountable for them? Second: what data does it depend on, and is that dependency lawful and ethical? Third: how will we know when it fails? Fourth: what is the human override? Fifth: what capability are we building so we are not permanently dependent on vendors?
Boards that institutionalise these questions build trust with regulators, customers and their own people.