Cloud-Native Identity And Access Models Tested In Large-Scale Studies

Cloud security research groups are running experiments on unified identity protection platforms that ingest logs from multiple clouds, SaaS applications, and on-premises directories to maintain a single, constantly updated risk score for each user and service principal. These projects explore how to blend login context, device health, privilege history, and data-access patterns so that conditional access rules can adapt in real time tightening controls when risk rises and relaxing friction when behavior looks routine. Investigators are also examining how such systems can reduce over-privileged accounts by supporting just-in-time access and time-bound elevation, while still giving security teams clear explanations of why a particular decision was taken instead of relying on opaque “black box” scores.​

12-12-2025