Having a similar issue in a cloud-based VM solution. Any solution that doesn't require a reboot? Like, keep the VMs live and pushin an MDM profile to it? This issue doesn't happen on baremetals, just on some VMs. Could this be caused by Apple's hypervisor framework?