Technical Anshu
Tutorials Blog
🔍 Policy Drift

Group Policy Drift After Copying Why Copied Settings Go Stale and How to Review Them

Every copied policy is perfect exactly once: on the day you import it. Then the months start editing. A technician relaxes a setting 'just for today', a Windows feature update adds two new policies nobody reviewed, and a fresh PC gets imaged from last year's backup. Drift is not a failure of the copy — it is what copies do over time. This guide is the review habit that keeps it visible.

✍️ Technical Anshu🪟 Windows 10 / 11⏱️ 7 minute read
🎥 Watch the full video — the GPO Transfer Tool is shown live on real PCs, step by step

Drift has three authors, and it helps to name them. People change settings locally and forget. Windows itself changes: feature updates ship new Administrative Template settings whose defaults may not match your intentions. And your own process drifts when a new PC is built from an old baseline backup instead of the current one. None of these show up as errors — the PCs just slowly stop being identical. The fix is not a tool you buy; it is a review rhythm you keep.

Drift spot-check — baseline v4 vs a suspect PC (example)
Baseline v4 (golden PC export) Accounts PC (LGPO /parse) USB storage ........ Blocked USB storage ........ Not configured Lock screen ....... 10 minutes Lock screen ....... 10 minutes Win Update ........ Ring 2 Win Update ........ Ring 1 Verdict: 2 settings drifted — re-apply baseline v4
👆 Example comparison — parse both policies to text and read the differences
1

Why a Copy Cannot Stay True by Itself

A domain GPO re-asserts itself at every refresh; a copied local policy asserts nothing — it simply sits where you left it while the world edits around it. Local administrators (including well-meaning support staff and 'optimising' utilities) can change any of it, and nothing phones home to tell you. Accept this and drift stops being a surprise: the question is never whether your copies drift, only how long they drift before you look.

2

The Quarterly Review — One Hour, Four Moves

  1. Re-export the golden PC as the new baseline version (even if you changed nothing — the export is dated proof of intent)
  2. Sample the fleet: pick five PCs at random plus any PC that had a support visit, and pull gpresult /h reports from each
  3. Compare: read the sampled reports against the baseline settings report — or export a suspect PC's local policy with LGPO and diff the parsed text
  4. Correct: re-apply the baseline to drifted PCs, and log what drifted and why — patterns in the log are tomorrow's policy changes or training gaps
3

When Windows Itself Moves the Goalposts

Each Windows feature update can add new policy settings and occasionally change what an old setting controls. After a fleet-wide version upgrade, open your baseline in the editor and walk the sections you care about — security, updates, privacy — reading for new entries marked with the new version. 'Not configured' on a brand-new setting means the Windows default rules, and the default is a product decision, not yours. Fold deliberate choices into the next baseline version rather than discovering them during an audit.

4

Version Your Backups Like Software

Drift thrives on ambiguity about which backup is current. Kill the ambiguity: every baseline export gets a version number and a date (Office Baseline v5 — 2026-10), a one-line changelog (what changed and why), and a single home — one folder on the share, one offline copy. Old versions stay, read-only, for rollback; they are never the folder a tired technician grabs at random. If two folders both claim to be 'latest', you do not have a baseline — you have a rumour.

5

Domains Drift Differently — Watch the Links

Domain fleets do not suffer local edits the same way — refresh re-asserts the GPO — but they drift at the management layer instead: GPOs edited directly instead of through change control, test GPOs left linked to production OUs, and 'temporary' security filtering that became permanent. The same quarterly hour covers it: export GPMC settings reports, review the link list per OU, and reconcile against your register. Different layer, same medicine.

🔴 The sentence that prevents drift: 'Policy changes happen on the golden PC (or in change control), never directly on a production machine.' Every drift horror story starts with someone breaking it just this once.

🎥 Watch the GPO Copy in the Full Video

See the policy backup leave one PC and land on the next — the complete transfer, done live.

▶️ Watch Full Video
❓ Frequently Asked Questions

How often should copied policies be reviewed?

Quarterly is the practical rhythm for a small fleet: re-export the golden baseline, sample a handful of PCs with gpresult reports, and re-apply where settings have drifted. After any Windows feature upgrade, do one extra review — new policy settings arrive with it.

Can I automatically detect drift on workgroup PCs?

There is no built-in free monitor for local policy drift. The workable pattern is scripted sampling: export a PC's local policy with LGPO, parse it to text with LGPO /parse, and compare against the parsed baseline. Anything beyond that is what management platforms — or a domain — are for.

A PC drifted because a technician fixed a real problem by changing policy. Now what?

Decide whether the change was right. If yes, promote it: make it on the golden PC, export a new baseline version, and roll it everywhere. If no, re-apply the baseline and record why. Drift logs exist so good fixes become policy instead of folklore.

This guide is based on the Technical Anshu GPO copy tutorial video • Stuck on a step? Ask in the video comments — you will get a reply 🙏