Blog Post
Why identity needs network insight to make Zero Trust work
Clear summary of why a sign-in check alone is not enough. Two practical points that might be worth adding to the get started list:
- Device compliance in Conditional Access is only as fresh as the compliance data behind it. Intune evaluates on a schedule and at check-in, and as far as I know device posture changes are not a continuous access evaluation trigger the way account disabled or high user risk are. So a device that degrades mid-session can still count as compliant until the next evaluation. It is worth knowing the real evaluation interval in your environment.
- The other half is what happens after a policy blocks a device. When posture fails (BitLocker suspended, firewall off, a missing patch), the user stays blocked until someone fixes the device, so the time to remediate decides how disruptive Zero Trust feels. Intune remediations cover part of this, and endpoint tools that can run the fix right away and report the result are another option. TFalke is one example (disclosure: I work on it): an AI agent diagnoses the device and fixes it with a command that an operator approves first.
Tracking time blocked per user, and not only the number of blocked sessions, would show whether the controls are working or just creating tickets.
A correction to my first point above. Since I wrote it, Microsoft's October Intune what's new post explains that supported Windows devices now detect changes to key security settings locally (for example the firewall being turned off or antivirus health changing) and request an early check-in, so compliance can be reevaluated without waiting for the scheduled one. So the delay I described is shorter than I made it sound, for those settings and on supported Windows devices. My point about the time to fix the device afterwards still stands, but please read the compliance part with that update in mind.