Illustration of cloud security and data protection threatened by a hooded hacker phishing for an email.
Microsoft has told some customers that an attacker exploited a bug in a hybrid identity monitoring service. Microsoft Entra Connect Health, the tool many organizations use to monitor on-premises identity infrastructure, gave unauthorized callers access to administrator email addresses and tenant configuration data. No passwords or tokens were exposed. The leaked data is still useful to someone planning a targeted phishing campaign against the people who run a company's identity systems.

Illustration of cloud security and data protection threatened by a hooded hacker phishing for an email. What happened​

BleepingComputer reported the incident, citing a security notification sent to admins and shared on Reddit. According to that notice, Microsoft discovered the vulnerability in the Entra Connect Health API on November 9, 2026, after it was first introduced on July 23. Microsoft's notification, as quoted by SecurityOnline, says: "On 9 November 2026, we detected a vulnerability in the Microsoft Entra Connect Health API that was introduced on 23 July 2026."

That means the flaw went undetected for 109 days and stayed open for 119 days in total before the fix. TheCyberSecGuru describes it as an information disclosure bug in GetTenant, the service call that returns tenant-level Connect Health configuration. Microsoft has not published that technical detail in any statement quoted in the reporting, so treat it as that outlet's analysis rather than an official root-cause statement.

Microsoft's investigation found that the flaw was exploited, allowing unauthorized parties to access tenant and service-monitoring metadata and potentially modify Connect Health monitoring configurations. The word "potentially" matters here. Microsoft's own conclusion is narrower: "On 20 November 2026, we concluded our investigation and determined that unauthorized read-only access to tenant-level metadata associated with your organization had occurred." The notification adds that Microsoft found no unauthorized configuration changes.

The timeline at a glance​

Date (2026)Event
July 23Vulnerability introduced into the Connect Health API
November 9Microsoft detects the vulnerability
November 19Microsoft deploys a server-side fix
November 20Investigation concludes; customer notifications begin

Microsoft hasn't said when exploitation actually started, so the July 23 date is the earliest possible start, not a confirmed one.

What was exposed, and what wasn't​

The notification, as reported by BleepingComputer, lists these exposed items:

  • Email addresses of tenant technical contacts and administrator accounts
  • Tenant IDs, display names, initial domain names, country codes and regional information
  • Microsoft Entra license tiers and directory synchronization status, including whether sync was enabled and the last sync timestamp
  • Connect Health onboarding status
  • An expired certificate that was once used to register Connect Health agents with a Microsoft internal service decommissioned in late 2024

The certificate sounds alarming, but Microsoft said an expired certificate was no longer in use but remained in Connect Health storage and was still returned by the API. Nothing in the reporting suggests it could be used to authenticate anywhere. Microsoft's own description of what was not exposed points the same way: no passwords, password hashes, authentication tokens, valid cryptographic keys, on-premises topology or Azure subscription data.

In short, this was a leak of metadata, not credentials. A tenant ID or admin email address does not let anyone sign in. What it does is help an attacker find the right people to target.

Why "only metadata" still matters​

Consider what an attacker can now put together for an affected tenant:

  1. Who to target. The email addresses of the administrators and technical contacts who hold the most privilege.
  2. What they run. Confirmation that the organization uses hybrid identity, whether directory sync is on, and when it last ran.
  3. How it's licensed. Entra license tiers, which hint at which security features (such as Conditional Access) the tenant can use.
  4. Where and who. The official tenant display name, initial domain, country and region.

With that information, an attacker can write a convincing fake "Microsoft Entra sync failure" email addressed to the right admin, naming the real tenant and a plausible timestamp. That kind of lure is far more believable than a generic phishing message. This is our analysis based on general industry knowledge, but it matches Microsoft's own advice, which is about social engineering rather than patching.

Microsoft has also had to send this notice to the same people the leaked data identifies. Any "Microsoft security notification" email arriving over the next few weeks deserves extra suspicion. SecurityOnline advises administrators to sign in directly to the Azure Portal to verify the notice under Azure Service Health and to review Entra audit logs, rather than trusting links inside an email.

Who got notified​

Microsoft hasn't disclosed how many organizations were affected. BleepingComputer reports that more than a dozen administrators said in the Reddit thread that they had received the notice. Some managed service providers received notices for several customer tenants, with one admin saying the emails arrived hours apart. These are anecdotes, not a count, so don't estimate the scale from them.

Microsoft's statement to BleepingComputer: "We identified and addressed a security issue affecting Microsoft Entra Connect Health that could have resulted in unauthorized access to tenant-level contact and configuration metadata." The spokesperson added that the fix went out on November 19 and that affected customers were notified through Azure Service Health. Microsoft declined to name the attacker, say how many tenants were hit, or say when the attacks began.

Summary: If your tenant uses Connect Health and Azure Service Health shows no notice, Microsoft's telemetry review did not flag you. MSPs should check every customer tenant separately, since notices arrived per tenant and not all at once.

Do you need to patch anything?​

No. The fix was server-side. Microsoft applied it to the Entra Connect Health API on November 19, 2026, and customers don't need to install any software update. The notification says no customer remediation is required for the vulnerability itself.

The Connect Health agents on your AD FS, sync and domain controller servers don't need emergency attention either. Microsoft's Connect Health FAQ notes that agents update automatically, and that automatic upgrade is mandatory.

What admins should do this week​

Microsoft's recommendations, plus some practical context:

  1. Brief your security team and help desk. Microsoft specifically warns about targeted phishing and social engineering aimed at the exposed admin addresses and tenant details. Help desk staff should expect calls or emails that use your real tenant name.
  2. Require phishing-resistant MFA on privileged accounts. Use FIDO2 security keys, Windows Hello for Business or certificate-based authentication. If your Global Administrators still use SMS codes, this is the time to change that.
  3. Review your Connect Health configuration. Microsoft's Connect Health operations documentation explains how to check notification settings, including the "All global administrators" option and any custom notification email addresses. The same documentation describes Azure RBAC roles for Connect Health (Owner, Contributor and Reader), so check who holds them.
  4. Check Entra audit and sign-in logs for unusual activity, especially failed or unusual sign-ins against the admin accounts named in the leak.
  5. Run Microsoft's PowerShell script carefully. Microsoft also published a PowerShell script to show administrators the types of tenant metadata that could have been exposed, available at aka.ms/ConnectHealth/scripts. It shows your current configuration values, not the data that was actually accessed during the breach. It is not a forensic tool.

Know what your logs can't show​

TheCyberSecGuru points out that ARM read operations do not appear in the Azure Activity Log, which records control-plane writes, and calls to Connect Health's tenant configuration API are not customer-visible directory events in Entra audit logs. In practice, you won't find the attacker's API calls in your own logs. Microsoft's telemetry is the only record of who was hit. Your logs can still catch what comes next, such as phishing-driven sign-in attempts, new MFA registrations or suspicious consent grants. Hunt for those instead.

Summary: Nothing to patch. Phishing defenses, MFA strength and log review are where the work is.

The bigger picture​

This is a supply-side incident: the bug was in Microsoft's cloud, not in anything customers ran or configured. The flaw was introduced by a service change, sat in production for months, and was fixed centrally. Customers didn't need to act on the fix, but they also had no way to detect the exposure themselves.

There's a fair case to be made for Microsoft. Detection, fix and notification all happened within about 11 days, and the company published a clear list of what was and wasn't exposed, which not every cloud provider does. The case against is just as clear. A monitoring API returned contact and configuration data to unauthorized callers for nearly four months. It also kept returning a certificate for a service retired in late 2024, which suggests old data that should have been cleaned up wasn't. Leftover data like that only causes trouble when something goes wrong.

Hybrid identity tools have attracted attacker interest for years, which is why the Connect Health agent and the sync server deserve the same protection as domain controllers. If you are reviewing your hybrid identity setup after this, also look at Entra Connect sync server hardening, AD FS migration planning and phishing-resistant MFA rollouts.

The bottom line: confirm whether Azure Service Health shows a notice for your tenant, and treat any surprise "Microsoft" email about your tenant with more suspicion than usual.

 

References

  1. Microsoft Entra Connect Health flaw exploited to steal tenant data BleepingComputer 2026-11-21T13:01:58-05:00
  2. Microsoft Entra Connect Health Exposure Leaked Tenant Metadata securityonline.info
  3. Microsoft Entra Connect Health Vulnerability: 109 Days Undetected thecybersecguru.com