Insights

When a Fake Teams Support Call Reaches Active Directory

Fake Microsoft Teams support calls can expose Active Directory weaknesses. Learn how attackers exploit them and what your IT provider should check.

Security analyst reviewing a protected workstation, identity certificates and internal servers, with the Plainsight Security logo on an illuminated panel.

Imagine a member of staff receiving a Teams call from someone who appears to be IT support. The caller says there is a problem with their computer and asks them to open Quick Assist. The employee enters the code they are given, and the caller can now control the machine.

No suspicious attachment was needed. The employee approved the remote session themselves.

In research published on 30 October 2026, Microsoft described two real intrusions that began this way. After gaining access, the attackers explored the organisation’s Windows network and looked for weaknesses in Active Directory Certificate Services (AD CS). The incidents show why the damage from a support scam depends on what an attacker can do after reaching a workstation.

A support call is only the first step

In both cases, attackers impersonated help desk staff through Teams messages and calls. They persuaded users to grant remote access, mainly through Quick Assist. From the affected workstation, they ran their own tools, checked the network and examined the organisation’s certificate services.

The attackers used legitimate communication and support tools to gain access through deception. The next stages depended on weaknesses in the internal environment.

AD CS helps organisations issue digital certificates to users and computers. Those certificates can be used to prove an identity when accessing other systems. If the service is configured too broadly, an attacker with ordinary access may be able to obtain a certificate representing a much more powerful account.

That is why a compromised workstation can become a domain-wide problem.

Two certificate weaknesses, two different outcomes

Microsoft observed two distinct routes.

An overly permissive certificate template (ESC1). In the first case, the attackers found a template that let a requester supply identity details and obtain a certificate suitable for authentication. Combined with broad enrolment rights and weak issuance controls, this allowed them to request certificates for privileged accounts. They used that access to perform DCSync, a technique that can retrieve sensitive domain credential data. Microsoft disrupted much of the activity, but one privileged account was excluded from automated containment. The attacker later used that account to access a file server and exfiltrate data.

NTLM relay to certificate web enrolment (ESC8). In the second case, the attackers induced a domain controller to authenticate and relayed that authentication to an AD CS web enrolment service. The service issued a certificate for the domain controller’s machine account. Microsoft reported that the customer isolated the associated workstation after alerts, and no further attack progression was observed.

The first weakness concerns who can request a certificate and which identity it can represent. The second concerns whether a web enrolment service will accept relayed authentication. Both can turn a foothold on an ordinary device into access to a powerful identity.

Why changing a password may not finish the job

A certificate issued to an attacker can remain usable independently of the password that first gave them access. Resetting that password or rebuilding the affected workstation does not, by itself, revoke the certificate. Tickets already issued through Kerberos also need separate investigation.

For an IT team responding to this kind of incident, the scope must include certificate requests and issued certificates, as well as the first user’s account and device. Microsoft’s investigation and mitigation guidance recommends checking certificate authority records, web enrolment logs, the identities involved and subsequent authentication activity.

Five questions to ask your IT provider

You do not need to know whether your network has an “ESC1” or “ESC8” issue to start a useful conversation. Ask:

  1. Do we run Active Directory Certificate Services, and do we know which certificate templates are available? In particular, are templates used for authentication restricted to the people and devices that need them?
  2. Can ordinary users request a certificate that represents another identity? Review requester-supplied identity information, enrolment permissions and whether approval is needed before a certificate is issued.
  3. Do we use AD CS web enrolment? If so, are those endpoints necessary, served over HTTPS and protected with Extended Protection for Authentication? Microsoft’s AD CS relay guidance explains the relevant safeguards. Changes to authentication settings should be planned and tested against existing applications.
  4. Who can start a remote support session with our staff? Is there a clear way to verify a support caller through a known internal channel before giving them control? Are external Teams contacts and the use of Quick Assist managed appropriately?
  5. Would we know if an unexpected certificate was issued? Confirm that certificate authority auditing is enabled, logs are retained and someone knows how to investigate and revoke an unauthorised certificate. Any already-issued authentication tickets need attention too.

If your organisation does not use AD CS, the certificate paths in these incidents do not apply to it. A verified support process and limits on remote access are still worth checking.

What internal testing adds

An external vulnerability scan can tell you a great deal about exposed systems and missing patches. It will not necessarily show whether a normal user can request a privileged certificate, whether a web enrolment endpoint accepts relayed authentication, or what an attacker can reach from a staff workstation.

An internal infrastructure penetration test can examine those relationships in context. The goal is to establish whether a realistic foothold could be turned into privileged access, then explain the changes that would break that path. Testing should be scoped and agreed with the organisation, particularly when checking sensitive services such as certificate authorities and domain controllers.

Microsoft’s investigation began with a convincing conversation, but the severity of the first incident came from what the attackers found behind it. That is the question worth asking of your own network: if someone gained control of one ordinary workstation, how far could they go?

If you would like an independent view, Plainsight Security’s internal infrastructure testing assesses Active Directory and the routes from a single foothold to more powerful access.

Portrait of Plainsight Security's lead tester

Written by

Mark Tomlinson

Our lead penetration tester, Mark Tomlinson, holds The Cyber Scheme Team Leader qualification in infrastructure penetration testing, an advanced certification recognised by the National Cyber Security Centre (NCSC) and used by professionals testing government systems and UK critical national infrastructure. Mark is also registered with the UK Cyber Security Council as a Principal Cyber Security Professional (PriCSP) specialising in Security Testing and holds an MSc in Computer Science with Cyber Security.

More about how we work
Talk to a tester

Put this into practice.

Cyber Essentials, Cyber Essentials Plus, and penetration testing — fixed-price, plain English, and built to stay out of your way.

← All insights