Explore our new Cybersecurity Terms Glossary, explained by our experts.

View the Glossary View the Glossary
PRIVIA

Cybersecurity Terms Glossary

What Are NTLM Coercion Techniques?

NTLM coercion forces a Windows server or domain controller to authenticate to a machine of the attacker’s choosing, supplying the authentication that a relay attack needs.

Illustration for the cybersecurity term NTLM Coercion
Home / Blog / Cybersecurity Terms / What Are NTLM Coercion Techniques?
17 November 2026

What Are NTLM Coercion Techniques?

What Are NTLM Coercion Techniques?

NTLM coercion is the art of making a Windows machine authenticate to you on demand. On its own it reveals nothing useful, but it supplies the missing ingredient for an NTLM relay attack: a victim that connects at the moment of your choosing, rather than whenever a user happens to mistype a share name. Pair coercion with relay, and an attacker with a single low-privileged domain account can often reach domain controller compromise in minutes.

The Underlying Behaviour

Several Microsoft RPC interfaces accept a UNC path as a parameter and instruct the server to go and look at it. When the server reaches out to \\\\attacker\\share, it authenticates using its own machine account over SMB or HTTP, because that is how Windows accesses network resources. The call is behaving exactly as documented. The problem is that the path is attacker-supplied and the authentication is automatic.

A domain controller’s machine account is a particularly valuable thing to coerce, because it is a member of Domain Controllers and in some configurations can be used to request a certificate or modify directory objects.

The Main Techniques

NameInterface abusedNote
PrinterBug / SpoolSampleMS-RPRN, RpcRemoteFindFirstPrinterChangeNotificationExPublished by Lee Christensen in 2018; requires the Print Spooler service to be running
PetitPotamMS-EFSRPC, EfsRpcOpenFileRaw and related callsReleased by Gilles Lionel in July 2021; originally worked unauthenticated against some systems
DFSCoerceMS-DFSNM, NetrDfsRemoveStdRootTargets the DFS Namespace service, which cannot simply be disabled on a domain controller
ShadowCoerceMS-FSRVP, the shadow copy serviceRequires the File Server VSS Agent role
WebDAV coercionAny of the above, with a WebDAV-style pathForces HTTP authentication instead of SMB, which bypasses SMB signing

The tool Coercer automates the whole set, trying every known method against a target until one responds. Expect new entries: this is a pattern in the protocol family rather than a finite list of bugs, and researchers have found fresh variants every year since 2018.

Why WebDAV Matters So Much

SMB signing, when enforced, prevents SMB relay. Coercion over WebDAV sidesteps that entirely by producing an HTTP authentication, and HTTP relay targets are plentiful. The most damaging is the Active Directory Certificate Services web enrolment interface. Relaying a domain controller’s coerced authentication to AD CS—the technique known as ESC8—yields a certificate for the domain controller’s machine account, which is then used to obtain a Kerberos ticket and perform DCSync. That chain, usually demonstrated as PetitPotam plus ESC8, remains one of the fastest paths from zero to domain compromise in an unhardened Active Directory.

The WebClient service that enables WebDAV is not running by default on servers, but it is on many workstations, and an attacker can often start it remotely through a searchConnector file dropped on a writable share.

Blocking It

Microsoft’s position has been that coercion is largely by design, so the defence focuses on making the resulting authentication useless rather than preventing the call.

Detection

Coercion is quiet but not invisible. Watch for machine account authentications arriving at hosts that are not file servers, especially from domain controllers. Certificate issuance to a machine account from an unusual requester is a high-fidelity signal for the ESC8 chain, and AD CS logs the requester identity. Network detection rules for the specific RPC operation numbers are also available in open Suricata and Zeek rule sets.

At Privia Security, we test this chain as standard during internal Penetration Test Services, because the gap between “patched” and “hardened” is where it lives.

FAQ – Frequently Asked Questions

Is NTLM Coercion a Vulnerability or a Feature?

Mostly a feature being used against its intent. The affected RPC calls are documented interfaces that legitimately instruct a server to access a path, and Microsoft has declined to treat several of them as vulnerabilities. Individual implementations have been patched where authentication was not required, but the broader behaviour persists, which is why defence focuses on making the coerced authentication unusable.

Does Disabling the Print Spooler Stop Coercion?

It stops PrinterBug only. PetitPotam uses the encrypting file system RPC interface, DFSCoerce uses the DFS namespace service, and others exist. Disabling the spooler on domain controllers is still worthwhile for other reasons, including PrintNightmare, but it should be treated as closing one door in a corridor rather than as a solution to coercion.

What Is the Single Most Effective Mitigation?

Enabling Extended Protection for Authentication and requiring HTTPS on Active Directory Certificate Services web enrolment, or removing that role if unused. Coercion only becomes critical when the authentication can be relayed somewhere valuable, and AD CS web enrolment is by far the most valuable target. Combine it with SMB signing and LDAP channel binding to close the remaining relay paths.

How Do We Know Whether We Are Exposed?

Run the public coercion tools from a standard domain-joined machine with an ordinary user account against your own servers, with authorisation, and see which respond. Then audit your AD CS configuration with Certipy or PSPKIAudit for ESC8 and related misconfigurations. The combination of a responding coercion method and an unprotected relay target is what turns theory into a domain compromise.

Related Terms

You May Be Interested In These

What Is a Business Impact Analysis?

Cybersecurity Terms

17 November 2026

What Is a Business Impact Analysis?

Read More
Why Are MD5 and SHA-1 Insecure?

Cybersecurity Terms

17 November 2026

Why Are MD5 and SHA-1 Insecure?

Read More
What Is the NIST AI RMF?

Cybersecurity Terms

17 November 2026

What Is the NIST AI RMF?

Read More