Microsoft confirms that it fixed an Entra Connect Health security issue and notified affected customers. Lawrence Abrams reports that attackers accessed administrator and technical-contact email addresses alongside tenant configuration metadata. The exposed details can help an impersonator make a message about an organization’s identity systems sound credible.
Entra Connect Health monitors an organization’s on-premises identity infrastructure through Microsoft’s cloud administration service. The reported vulnerability was in that cloud service’s API. Microsoft says its November 19 service-side fix closes the flaw, with no customer software update required to repair it.
Evidence status: Microsoft-confirmed security incident, with unauthorized reads described in its customer notification. Abrams obtained a spokesperson’s confirmation. The full notification reproduced by an administrator supplies additional detail, but remains a customer-posted copy.
Potential changes and observed access are different findings
The notification says Microsoft detected the vulnerability on November 9, after its introduction on July 23. It says the investigation concluded on November 20 and customer notices followed. Those dates describe discovery, repair and investigation, rather than a verified start date for attacks.
The vulnerable API could expose monitoring metadata and potentially allow changes to Connect Health monitoring configuration. For the notified organization, Microsoft’s copied notice describes unauthorized read-only access and says investigators identified no unauthorized configuration changes. Potential write capability should stay separate from the observed activity.
The reported exposed fields include:
- Administrator and technical-contact email addresses. These identify useful targets for phishing and support impersonation.
- Tenant identity and region information. Tenant ID, display name, initial domain, country code and regional scope were included.
- Service configuration metadata. The notice lists Entra license tier, directory synchronization status and last synchronization time, plus Connect Health onboarding status.
- A legacy, expired certificate. Microsoft says it belonged to an internal registration service decommissioned in late 2024 and was no longer used, although the API still returned it.
Microsoft says this vulnerability excluded passwords, password hashes, authentication tokens, valid cryptographic keys, on-premises server and Active Directory topology details, and Azure subscription information. The reviewed evidence supports contact and configuration exposure, not a claim that attackers obtained every user’s credentials.
Verify the notice through Azure Service Health
Affected-tenant scope matters. Administrators should open the Azure Portal directly and check the Azure Service Health security advisory for their tenant. A copied email or another administrator’s experience cannot establish whether a particular organization was affected.
Microsoft’s notice recommends these precautions:
- Brief security staff and technical contacts. Watch for messages invoking real tenant details, Connect Health problems or urgent account repairs. Verify requests through established channels.
- Check privileged-account protection. Require phishing-resistant multifactor authentication, such as security keys or appropriately configured passkeys. Review just-in-time privileged access where applicable.
- Review Connect Health configuration and relevant logs. Compare settings with expectations and examine Entra audit and sign-in logs and Azure Activity Logs for unusual administrative activity.
- Understand what the metadata helper can show. Microsoft links a reference-script repository. Its
GetTenantquery returns current configuration, which can differ from historical exposed values. It does not reconstruct the attacker’s access history.
The repository documentation describes several other queries as well. Their presence is not evidence that all those datasets were exposed in this incident. Follow the advisory applicable to the tenant rather than expanding the breach from a directory listing.
Caveat: Microsoft has not publicly disclosed the attacker, affected-tenant total or first exploitation date; the detailed exposure account relies on its investigation and customer notification.