Audit logs hold lots of information, including records for when Azure AD consent permission grants happen. Checking the audit data can detect illicit grants. Records are in the Azure AD audit log and are also ingested into the Office 365 (unified) audit log, so there’s two places to check. The audit data is interesting and could help administrators work out if a permission grant is illicit. But only if checks are made and people review the reports.
The preview of a new app governance add-on for Microsoft Client App Security gives Office 365 administrators insight into Graph-based apps. The add-on depends on information gathered from Azure AD and MCAS to generate insights about apps and their usage, including highlighting apps which are overprivileged or highly privileged. Although you can do some of the auditing yourself, the add-on makes it easier. It’s a preview, so some glitches are present.
The Office 365 audit log is packed full of information about what happens inside workloads. New events show up all the time. The question is how to understand what actions these events relate to. We outline a simple procedure to discover the presence of new audit events and dive into the investigation of an event called Consent to application, which is pretty important in the context of recent high-profile attacks.