Windows Hardening & Continuous Compliance: Keeping 200+ Security Settings Secure 24/7

Every Operating System – not even counting the software installed on top of it – has a myriad of configuration settings that affect security. Windows Server alone has 4000+ configurable options, out of which 200+ are important for security & compliance efforts. They include:

  • SMB, RDP, UAC, …
  • PowerShell (execution policy, script block logging, transcription)
  • Various Logging Options (audit policies, log sizes, retention)
  • User Rights Assignments
  • Network settings like LLMNR, NetBIOS over TCP/IP and LDAP signing

These settings are at the very core of all of your security & compliance efforts – and they are surprisingly often ignored. Leaving the very center of your end points unprotected makes it that much more difficult for other security tools to do their job – whether it’s AV, EDR, a firewall, and the list goes on. A single wrong setting will not only cause compliance risks – but allow an attacker (internal or external) to compromise a host or an entire AD domain.

Why a single setting matters

Point in case: LLMNR is enabled out of the box on every Windows installation, and all it takes is one mistyped share name on a workstation for a tool like Responder to capture a NTLMv2 hash – or relay it straight to another host if SMB signing isn’t required there either. The same goes for SMBv1 that was never disabled, a service account that somehow ended up with “Act as part of the operating system” or “Debug programs”, or PowerShell script block logging that was never turned on, which simply means that whatever an attacker runs in PowerShell will never show up in any log. None of these will be flagged by a vulnerability scanner, since there is no CVE to flag – yet every single one of them is a well-traveled path to Domain Admin.

Making sure that your Windows servers and workstations are hardened is not only a key requirement for compliance – it also reduces your attack surface for both internal and external threats. This isn’t just our opinion either – essentially every framework and benchmark out there boils down to these settings in one way or another. The CIS Benchmarks for Windows Server and Windows 11 are, at their core, a list of several hundred recommended values for exactly these settings, the DISA STIGs cover the same ground for DoD environments, and NIST 800-171 (and with it CMMC) explicitly requires that you establish a secure baseline configuration (CM.L2-3.4.1 & 3.4.2) and then keep it that way. PCI DSS, HIPAA and ISO 27001 phrase it differently, but they all end up in the same place: Know what your settings are supposed to be, make sure they are, and be able to prove it to an auditor.

But how do you make sure that all these settings not only are correct (= secure), but also stay correct, secure and compliant?

Why Group Policy & point-in-time audits aren’t enough

Most organizations configure everything correctly once, usually with Group Policy (reminder: you can monitor group policy changes with EventSentry), and then hope and/or assume that it stays that way. Alas, settings drift: A vendor installer re-enables a service, a new GPO with a higher link order overrides an older one, a feature update resets a registry value, or an admin “temporarily” relaxes UAC on a server to troubleshoot something and never puts it back. Group Policy is an enforcement mechanism, not a verification mechanism – it re-applies most settings only when the GPO itself changes, so a value that was modified locally can stick around for quite a while, it doesn’t apply to workgroup machines and cloud VMs that were never domain-joined. A quarterly audit or a one-time hardening project tells you that a host was compliant on that day – but not the next day and beyond.

Vulnerability scanning is important and should be part of any comprehensive security strategy, but it can’t evaluate the majority of security and configuration settings in Windows – only software that runs locally (read: an agent) can evaluate these settings.

What is continuous compliance?

Enter continuous compliance, where all relevant security and configuration settings are constantly evaluated and verified. Automatically. No scans to schedule or run, no machines to connect to – just up-to-date validation of all monitored hosts, no matter if it’s a local DC or a laptop at someone’s house. In other words, instead of finding out during the next audit (or, worse, during the next incident) that a setting has been wrong for six months, you find out when it changes – and you can fix it before anybody else notices.

Continuous compliance with EventSentry

Continuous compliance is made possible by EventSentry’s validation scripts, a module which contains over 200 checks that continuously run in the background on every monitored host. Results are reviewed in the web reports and can be classified by compliance framework (e.g. CMMC, NIST 800-171), severity, host and more, which also makes for a pretty convenient piece of evidence when the auditor asks how you monitor your baseline. Remediation steps are provided for every single script, with group policy remediation coming in v6.1.

Chances are that at least one of those 200+ settings on your network isn’t what you think it is – the question is just whether you find out before or after somebody else does.