Guides

Plan backups that can survive ransomware damage

Separate backup copies and access, check version retention and deletion controls, practice restoration, and prepare a recovery plan for ransomware damage.

A backup plan for ransomware needs to answer more than “Are my files copied?” Ask whether an attacker who controls the everyday computer or its accounts could also erase those copies, replace every older version, or prevent you from accessing them. Then establish how you would restore wanted data onto a suitable device.

This guide is for prevention and readiness on home or small-office systems. Managed business systems need an administrator-led design and incident plan. If an attack may already be changing your files, leave preserved backup storage disconnected and start urgent triage. Do not connect every drive to check whether one still works. Ransomware explained describes the threat; pre-cleanup preservation covers suspect current data.

Write a recovery goal you can check

Name the information you need first and the latest acceptable point you could recover to. A student may prioritize a current project; a small office may need customer records and the application that reads them. A readable photo copy does not establish that a business database can be recovered.

Include the dependencies: the recovery device, backup software, export format, account access, encryption keys, and instructions. Keep an accessible private copy of the plan outside the original computer. For managed systems, name the responsible person and approved support process rather than improvising a separate personal backup arrangement.

Choose a routine whose gaps you understand. A weekly offline copy can preserve an earlier state, but changes since that copy are not automatically included. Consider both how much recent work could be lost and how long it would take to find, transfer, and use the retained data. These are planning questions, not a guaranteed recovery time.

Separate the copies and the ways to change them

The NCSC’s offline-backup discussion describes the familiar 3-2-1 pattern: at least three copies, on two devices, with one offsite. It also explains that this pattern alone does not require an offline copy. Counting three copies is therefore a starting inventory, not proof they resist the same attacker.

For a simple personal setup, consider the working files, a periodically disconnected backup, and another appropriately protected copy in a separate location. Check each copy’s versions and access independently. Two USB drives connected together during every backup still share a period of exposure. A cloud copy writable through the compromised account may share the same destructive access even though it is geographically distant.

Use the offline backup guide for a practical connect, verify, eject, and store routine. This arrangement has limits: the computer can affect a writable drive while connected, and the offline version may already contain unwanted content from an earlier incident. Retain history and test retrieval instead of calling a disconnected drive automatically clean.

Understand which protections solve which problem

Backup terms describe different controls. Evaluate the actual configuration rather than a product badge:

ControlQuestion it helps answerLimitation to investigate
Offline copyCan the ordinary device reach this copy now?It is exposed while connected and may be stale
Offsite copyWould loss of the original location destroy every copy?Remote location does not remove account-based access
Version historyCan I choose a state before unwanted changes?Versions expire or may be removable by an attacker
Encryption at restCan someone read stored contents without the key?An unlocked copy can still be changed or deleted
Immutable or locked retentionCan retained data be altered or removed during the defined period?Scope, policy changes, and privileged access vary
Restore exerciseCan I retrieve and use the selected data through this process?One successful sample does not validate everything

NCSC’s ransomware-resistant backup principles are aimed at vendors and organizational system owners. They distinguish resistance to destructive changes from protection against stolen-data extortion. A usable backup can reduce data-loss consequences; it cannot undo confidential information already copied by an attacker.

“Immutable” needs a specific explanation from the provider: which copies are protected, for how long, who can change the setting, and whether the protection applies to existing data. Do not assume a recycle bin or an ordinary read-only user view provides that property. Avoid enabling irreversible retention settings without understanding their storage and administration effects.

Check history, deletion rights, and recovery access

Use the provider’s current documentation for your account, edition, and policy. Ask whether older states remain available after mass changes, deletion, a full storage quota, or a lost account. A backup that retains only a small number of runs may replace every useful earlier state after several bad runs. Determine whether retention is based on elapsed time, copy count, capacity, or a combination.

NCSC’s on-premises principles discuss isolated administration, separate credentials, protected keys, and retained earlier states. For an administrator, that means evaluating who can manage or destroy the backup—not merely who can write working files. For a home user, start by identifying which saved account can change each copy and keeping recovery credentials accessible from a trusted alternative device.

Its cloud principles address destructive requests, customer lockout, version retention, and alerts. Ask the provider how you recover if ordinary sign-in or company identity systems are unavailable. Use supported additional sign-in protection and independent recovery information, following your organization’s policy.

Do not rely on signing out of one browser to revoke every backup client’s access. Similarly, renaming a network share does not establish isolation. For a shared server or network-attached storage (NAS) system, have the administrator review the management interface, writable shares, maintenance updates, and restore permissions.

Ordinary cloud synchronization has a useful availability role, but unwanted changes may propagate. The file restoration guide explains OneDrive’s distinct version, deletion, and rollback limits. Confirm your own account’s recovery features; a feature described for one plan is not automatically present in another.

Watch the routine and rehearse a bounded restore

Keep a small record of backup completion, exceptions, available space, retention changes, and the latest successful retrieval test. Assign someone to notice failures. An alert delivered only to an inaccessible mailbox is less useful when that account fails.

For a readiness exercise, retrieve known ordinary sample data into a separate destination using the documented process. Include an older version and a representative important application format where practical. Verify that the required keys, account route, and application are available without depending entirely on the original device. Record the scope and result honestly.

Do not run unknown executables or restore an entire suspect browser profile merely to prove that a backup works. A file-list comparison, a readable personal document, and a full application recovery answer different questions. For business continuity, an administrator should plan a safe representative exercise with the required dependencies.

Review the arrangement after changing devices, accounts, storage, or important applications. Check whether the old copy is still accessible before retiring it. Avoid allowing repeated failed jobs or space-pressure pruning to silently shrink the history you expected.

Keep incident recovery separate from routine backup

If files are already changing or a ransomware finding is present, use the incident investigation before reconnecting a recovery copy. Its containment and preservation decisions belong ahead of a routine backup exercise.

During suspected ransomware damage, stop ordinary use of the affected system and seek appropriate incident help. Protect preserved copies and retain relevant details. An immediate new copy of damaged files is preservation for investigation; it does not replace an intact earlier version.

NCSC’s malware mitigation guidance advises connecting backups only to suitably clean recovery devices and checking for malware before restoring. A scan contributes evidence but does not certify every retained item. The incident may predate the visible damage, so the newest backup is not always the right version.

Use post-malware verification to assess the destination and selective file restoration for source, version, and overwrite decisions. For cloud restoration, account for other connected devices that could repeat unwanted changes. Address exposed credentials through account recovery separately from file recovery.

The outcome to aim for is a documented, retrievable set of wanted data and a known restoration path. It is not a claim that ransomware can no longer cause harm. This article is researched guidance; no ransomware, backup-product effectiveness, or recovery-time experiment was performed.

Sources & documentation

Sources checked on 6 Oct 2026. Documentation research does not establish independent testing.

MalwareCleanupTools

Guides based on official documentation. About our content · Corrections policy