New white paper on applying the methodology of zero trust to reduce AI making it up and doing the wrong thing.