A risk-based approach to mitigating the Gen(AI) challenge to assessment integrity