The Trump administration's framework for assessing potential cybersecurity risks posed by advanced AI reportedly has no interest in testing open models. Axios reports that not only do the voluntary gu...