Control. Use approved devices and infrastructure as part of your AI capacity.. Treat approved devices and compute resources as part of the workspace so suitable AI work can run where capacity and policy allow.. Fleet Intelligence. Enrol approved devices or compute nodes. Route suitable workloads across available capacity. Keep sensitivity and network policy attached to each workload. AI routing. Private AI. Deployment. Secure execution. AI routing can prefer device, private or cloud capacity depending on policy and availability.. Ownership, network policy, sensitivity rules and audit history stay attached to every node and workload.