Sunday, November 15, 2026

NoID Privacy for reversible Windows 11 hardening and telemetry reduction

NoID Privacy for reversible Windows 11 hardening and telemetry reduction

NoID Privacy applies a 600+ check Microsoft Security Baseline-derived hardening profile to Windows 11 24H2-26H2 via a backup-apply-verify-restore PowerShell engine for authorized administrators and power users.

ToolNexusOne23/noid-privacy — PowerShell security and privacy hardening tool for Windows 11 24H2/25H2/26H2, version 2.3.1, GPL-3.0
CategoryWindows system hardening / privacy configuration (PowerShell)
Primary UseApplying and verifying a Microsoft Security Baseline-derived profile with ASR rules, Defender, Edge, DNS-over-HTTPS, telemetry and AI policy hardening on standalone Windows 11 machines
Safe UseHardening your own standalone systems, VMs, test/dev and air-gapped machines in authorized administrative contexts; not for production domain-joined systems without AD team coordination
Telemetry NoteThe tool reduces Microsoft telemetry rather than generating it; all changes are written to the local Registry, sealed in prestate snapshots, and Verify-Complete-Hardening.ps1 produces a local HTML report defenders can audit

NoID Privacy is a Windows PowerShell 5.1 hardening engine from NexusOne23/noid-privacy that applies more than 600 declared configuration checks across seven modules on Windows 11 24H2, 25H2, and 26H2 x64 clients. Its defining architectural idea is BAVR — Backup, Apply, Verify, Restore — meaning every registry value and policy state it touches is recorded before mutation and can be put back exactly. The current release is version 2.3.1 under GPL-3.0, with 158 stars and an active commercial sibling, NoID Privacy Pro, that wraps the same open-source engine in a guided Windows app. This is a defensive configuration tool, squarely in the tradition of baseline hardening scripts, and it is aimed at professionals and small businesses that manage machines without Intune or Active Directory.

The security foundation is a 425-target profile derived from the Microsoft Security Baseline for Windows 11 26H2, safety-reviewed for the fully supported client builds. On top of that sit 24 Microsoft Edge v151 baseline values plus 9 separately labelled privacy additions, 19 Attack Surface Reduction rules, and policy configuration for VBS and Credential Guard where hardware and licensing support it. The README is candid that Credential Guard entitlement is limited to Enterprise and Education editions per Microsoft's own documentation. This is a thoughtful split: the baseline layer tracks Microsoft's guidance, while the privacy layer is clearly labelled as additions beyond it, so an operator always knows which decisions are Redmond's and which are the tool author's.

The privacy layer is where NoID Privacy differentiates itself from a pure baseline script. DNS hardening lets you choose a documented public resolver with Windows DNS-over-HTTPS enforcement, with filtering behavior dependent on the provider selected. Telemetry configuration ships in three modes — MSRecommended, Strict, and Paranoid — giving operators a graduated dial rather than a single on/off switch. The AntiAI module is particularly interesting for its era: 12 AI policy groups with typed registry and URI restoration, plus handling of Windows Recall snapshots and component removal. The README is explicit that Recall snapshot deletion crosses a separate recovery boundary — deleted snapshots cannot be restored by the BAVR mechanism.

Bloatware removal is treated with unusual rigor for this tool category. It is two-tier and explicit about its destructive boundary: policy and registry state restores exactly, and sealed Tier 1/Tier 2 app identities allow separate original-user package re-registration with a verified Store fallback — but deleted application data is unrecoverable. This kind of honest disclosure about what is reversible and what is not deserves mention, because many consumer privacy scripts silently conflate the two. An operator reading the README knows precisely which mutations are sealed prestate and which are permanent before pulling the trigger.

The installer deserves its own analysis because bootstrap security is where hardening scripts most often go wrong. The reviewed bootstrap downloads install.ps1 from an exact tagged release rather than piping network content into execution, and the README instructs users to inspect the local file with Get-Content before running it. The installer then resolves the tagged release, requires the exact ZIP plus its CHECKSUMS.sha256, validates archive paths, types and resource bounds before extraction, and validates version, syntax and JSON in same-volume staging before an atomic swap with rollback protection. It fails closed if GitHub is unavailable, checksums mismatch, or staging validation fails, and never falls back to an unverified main-branch archive. From 2.3.0 onward the release scripts are Authenticode-signed, verifiable via Get-AuthenticodeSignature.

Installation in two flavors: the reviewed bootstrap fetched with Invoke-WebRequest to a temp file, inspected, then run under a process-scoped -ExecutionPolicy Bypass; or the manual route of git clone https://github.com/NexusOne23/noid-privacy.git followed by Start-NoIDPrivacy.bat as Administrator. The README's handling of execution policy is technically literate: it explains that Windows 11 ships with the Restricted policy, that browser-downloaded files carry the mark-of-the-web while Invoke-WebRequest output does not, and that Group Policy-set execution policies override process-scoped values. On shared or managed PCs it recommends -InstallPath "$env:ProgramFiles\NoIDPrivacy" so only administrators can modify scripts that run elevated — a small detail that reveals a security-first authoring mindset.

The BAVR pattern itself is the tool's core contribution. Rather than a full-machine backup, it seals the exact prestate of its declared mutation targets — a scoped, auditable snapshot tied to the settings the tool actually changes. The README is careful to state this is not a substitute for a system image, and mandates a restore point, full backup, or VM snapshot before running. This separation between tool-level reversibility and system-level recovery is the correct engineering framing, and it distinguishes NoID Privacy from the many one-shot debloat scripts that leave users stranded when something breaks.

Verification is first-class rather than an afterthought. The Verify-Complete-Hardening.ps1 script can be run standalone with an -Assessment flag for a pre-hardening check that changes nothing, or from the interactive menu's [V] option after a reboot. Results land in an HTML report available in ten languages. For defenders and auditors, this is the most valuable artifact: a documented, explicit verification of which of the 600+ declared targets are actually in effect, which can serve as evidence in a compliance or CIS-style review workflow on standalone machines.

The domain-joined warning is prominent and correct: the tool writes effective local policy and does not create or edit AD Group Policy objects, so domain GPOs can silently overwrite local hardening during startup, sign-in, or background refresh. The README recommends standalone systems, home PCs, VMs, air-gapped machines, and test/dev environments, and explicitly tells enterprise users to integrate the settings into Domain Group Policies instead. This is honest scope definition — the tool knows exactly where it belongs and where it will lose fights to gpupdate.

The commercial layer is worth noting for transparency. NoID Privacy Pro (from €39 one-time) is the same engine with a Windows app, three guided profiles (Quick & Secure, Balanced, High Security), one-click actions each with their own backup, and in-app visual reporting; its free check runs without a license and compares a machine against the Quick & Secure profile without changing anything. The Pro binary validates its license online at startup and checks for updates at most every 24 hours, sending no usage data. Everything in the GitHub repository remains fully functional and free under GPL-3.0, with purchases funding the open-source engine — a sustainable model that does not cripple the free edition.

Where does this fit in an authorized workflow? Think baseline enforcement on standalone workstations, air-gapped analysis VMs, small-business machines without management infrastructure, and test environments where you want a reproducible, verifiable, reversible hardening state. The telemetry modes and DoH enforcement also make it useful for privacy-sensitive research environments where you want to minimize and control outbound Microsoft traffic. The 26H2-derived baseline means it tracks current Microsoft guidance rather than stale CIS-style mappings, and the seven-module structure means you can reason about which domains of configuration you are touching.

Caveats for operators: the tool modifies the Windows Registry and system state, so treat it as a privileged configuration change with a change-control mindset — read the module list, run the -Assessment first, snapshot, then apply. The checksum file protects against corruption and mismatched downloads but is not a code signature, so the Authenticode check on extracted scripts and independent review of the bootstrap remain the real trust anchors. With 158 stars, a disciplined release process, signed scripts, and a SECURITY.md documenting the trust boundaries, NoID Privacy sits comfortably in the upper tier of Windows hardening tooling — worth evaluating for any authorized standalone Windows 11 estate.

Official project repository for NexusOne23/noid-privacy.
Download Tool

Educational analysis for authorized security professionals. Use only in controlled, authorized environments.

Sponsored
Share articleFacebookXLinkedIn

Continue exploring

Browse all articles →

0 comentários:

Post a Comment

Note: Only a member of this blog may post a comment.