Insecure Setting of Generative AI/ML Model Inference Parameters

The product has a component that relies on a

generative AI/ML model configured with inference parameters that

produce an unacceptably high rate of erroneous or unexpected

outputs.