Join our Newsletter — 33% off our NHI Course
Home› Guides› Backup and Recovery Policy Template: Ransomware-Resistant Backups With…
Template Security Policies & Awareness

Backup and Recovery Policy Template: Ransomware-Resistant Backups With Tested Restores

PDF for newsletter subscribers · free to read online

← All guides
By NHI Mgmt Group, edited by Lalit Choda Updated 27 October 2026 12 min read
On this page

Ransomware operators look for backups first, and most organisations discover during an incident that their backups were reachable with the same administrator credentials, were never restored end to end, or did not include the SaaS data the business now runs on. This template sets the rules for what is backed up, how often, how long it is kept, how backups are isolated and protected from deletion, how backup credentials and keys are governed, and how restores are proven by testing. It is written for CISOs, infrastructure and cloud platform owners and business continuity leads; have legal and compliance review retention periods before adoption. Recovery procedures for a declared disaster belong in the Disaster Recovery Plan Template; this policy governs the backups that plan depends on.

How to use this template

  • Replace text in square brackets, such as [Organisation], [Head of Infrastructure] and [35 days], with your own terms and values. Bracketed values are suggested defaults, not external requirements.
  • "Must" indicates a mandatory requirement. "Should" indicates a recommended practice where a documented exception may be justified.
  • Recovery point and recovery time objectives in section 4.1 must come from your business impact analysis, not from this template. See the Business Continuity Plan Template.
  • Keep the backup schedule (section 4.2) as a separately versioned standard so platform teams can change it without re-approving the policy.
  • Place this policy beneath your top-level information security policy.

1. Purpose

This policy sets the requirements for backing up [Organisation]'s information and systems so that they can be restored to a known good state within agreed time objectives after deletion, corruption, hardware failure, a cyber attack such as ransomware, or loss of a site or cloud provider. Its aim is that every critical service has a backup that an attacker with administrator access to production cannot delete, encrypt or alter, and that restores are proven by testing rather than assumed.

2. Scope

This policy applies to all information owned, processed or held by or on behalf of [Organisation], including:

  • Servers, virtual machines, hypervisors, databases, file shares and storage in [Organisation] data centres and cloud accounts.
  • SaaS applications holding [Organisation] data, including email, collaboration, CRM and source code platforms. A provider's own resilience is not a backup of [Organisation]'s data.
  • Identity infrastructure: directories, identity provider configuration, privileged access vaults, certificate authorities and key management configuration.
  • Infrastructure-as-code templates, configuration, golden images and the source code and executables needed to rebuild systems.
  • Endpoints only where they hold data not stored in a managed service (for example [named engineering or OT workstations]).

It applies to all staff, contractors and service providers who operate systems or backup services for [Organisation].

3. Roles and responsibilities

RoleResponsibilities
[Head of Infrastructure]Owns this policy and the backup standard; operates the backup platforms; reports backup and restore test results.
[CISO]Approves isolation, immutability and credential requirements; approves exceptions for Tier 1 services; reviews backup security alerts.
System and data ownersClassify their systems into a recovery tier; confirm what must be backed up; take part in restore tests and sign off results.
[Backup administrators]Configure jobs, retention and immutability; monitor job success; perform restores. Hold dedicated backup accounts, not production administrator accounts.
[Business continuity lead]Provides recovery objectives from the business impact analysis; schedules recovery exercises.
[Security operations]Monitors backup platform logs and alerts; treats tampering with backups as a security incident.
[Data protection officer]Advises on retention of personal data in backups and on handling erasure requests.

4. Policy statements

4.1 Recovery tiers and objectives

TierExamplesRecovery point objective (RPO)Recovery time objective (RTO)Isolated or immutable copy
Tier 1: CriticalIdentity provider and directory, payments, customer-facing core services, [named systems][15 minutes to 4 hours][4 to 24 hours]Must
Tier 2: ImportantInternal business applications, collaboration, source code[24 hours][2 to 3 days]Must
Tier 3: StandardFile shares, departmental tools[24 hours][5 to 10 days]Should
Tier 4: RebuildableStateless workloads rebuilt from code and imagesNot applicable: back up the code, images and configuration[As rebuild allows]The code and images must be
  1. Every system must be assigned a tier by its owner and recorded in [the asset inventory or CMDB]. A system with no tier is treated as Tier 2 until classified.
  2. RTO must include the time to rebuild the platform the data is restored to, not only the time to copy data.

4.2 Backup copies and schedule

  1. Tier 1 and Tier 2 data must have at least three copies (production plus two backups), on at least two different storage systems or providers, with at least one copy that is offline, logically air-gapped or immutable, and at least one copy in a separate location from production.
  2. Backup frequency must meet the RPO for the tier. Databases in Tier 1 must use transaction log or continuous backup where the RPO is under [4 hours].
  3. Backups must include the configuration needed to restore: identity provider and directory configuration, DNS, firewall and network configuration, KMS key policies, and infrastructure-as-code. CISA's #StopRansomware Guide recommends keeping offline backups of infrastructure-as-code templates and golden images of critical systems.
  4. SaaS data in scope must be backed up with a tool or export that stores copies outside the SaaS tenant, under [Organisation]'s control.

4.3 Retention

Backup typeDefault retention
Daily[35 days]
Weekly[13 weeks]
Monthly[13 months]
Annual or legal holdAs set by [the retention schedule] and legal hold notices
  1. Retention must be based on time periods, not on a fixed number of backup versions, so a burst of corrupted backups cannot push the last good copies out of retention. This follows the NCSC principles for ransomware-resistant cloud backups.
  2. Immutable retention must be at least as long as [the expected time to detect an intrusion, for example 30 days], so that a clean restore point exists from before the attacker arrived.
  3. Personal data in backups must not be kept longer than the retention schedule allows. Where an erasure request cannot be applied to a backup, the data must be put beyond use and removed again if the backup is restored. See the Data Retention and Disposal Policy Template.

4.4 Isolation and immutability

  1. Backups of Tier 1 and Tier 2 data must be protected against deletion or alteration for their retention period, using write-once storage, object lock, vault lock, offline media or an equivalent control that cannot be overridden by a single administrator.
  2. Deletion of backups, reduction of retention and disabling of immutability must require [two-person approval] through a separate channel, and should be delayed by [a soft-delete period of at least 14 days] so alerts can trigger a response.
  3. The backup platform's management interface must not be reachable from the general corporate network. It must be on a separate management network or accessible only through [the privileged access workstation or PAM solution].
  4. At least one copy of Tier 1 data must be in a separate security domain: a different cloud account, subscription or tenant, or a different provider, with no trust relationship that lets production administrators manage it.
  5. [Organisation] must retain an out-of-band way to reach backups if corporate identity and IT are unavailable, such as a break-glass account held offline.

4.5 Backup identities, credentials and keys

Backup systems run on powerful non-human identities: service accounts and API credentials that can read every system they protect. They are among the first credentials an attacker seeks.

  1. Backup administrator accounts must be separate from production and domain administrator accounts, must not be joined to the production directory where the platform allows, and must use phishing-resistant MFA.
  2. Service accounts, API keys and application secrets used by backup software must be inventoried with a named owner, scoped to read only what they back up, stored in [approved secrets manager], and rotated at least [every 90 days] or on staff change. See the Backup Identity Isolation Guide.
  3. Backups of Confidential and Restricted data must be encrypted. Encryption keys must not be stored with the backups they protect and must themselves be backed up, including an offline copy of master keys, so loss or deletion of a key cannot make backups unreadable.
  4. Credentials that can restore data into production must be held in [PAM], checked out with approval, and their use logged.

4.6 Restore testing

TestTier 1Tier 2Tier 3
Automated backup integrity checkEvery jobEvery jobEvery job
File or record restore sample[Monthly][Quarterly][Annually]
Full system restore to an isolated environment, timed against RTO[Quarterly][Annually][Every 2 years]
Restore from the isolated or immutable copy, without production credentials[Annually][Annually]Not required
Recovery exercise including identity provider rebuild[Annually]As part of the Tier 1 exercise where dependent
  1. Restored data must be checked by the system owner for completeness and integrity, and for signs of compromise, before it is used in production. NIST CSF 2.0 RC.RP-03 requires the integrity of backups to be verified before restoration.
  2. Failed tests must be raised as [high priority] findings and retested within [30 days].

5. Exceptions

  1. Systems that cannot meet this policy (for example legacy or OT systems without supported backup agents) must be recorded in [the exception register] with compensating controls, such as image-level backup or manual export, and a review date.
  2. Exceptions for Tier 1 systems are approved by [CISO]; others by [Head of Infrastructure]. Exceptions expire after no more than [12 months].

6. Compliance and monitoring

  1. Backup platforms must forward logs to [the SIEM]. Alerts must cover, whether or not the action succeeds: deletion of backups or restore points, changes to retention or immutability, disabled or failed jobs, changes to backup administrator accounts and roles, new API keys, encryption key changes, and restores to unusual destinations.
  2. Backup job failures for Tier 1 systems must be investigated within [one business day].
  3. The [Head of Infrastructure] must report quarterly on:
    • Percentage of in-scope systems with a tier and a backup meeting it.
    • Backup job success rate by tier.
    • Tier 1 and Tier 2 systems with a current immutable or isolated copy.
    • Restore tests completed, passed, and the measured restore time against RTO.
    • Backup service accounts and keys with a named owner and in-date rotation.
  4. Compliance with this policy is subject to internal audit at least [annually].

7. Review and approval

ItemValue
Policy owner[Head of Infrastructure]
Approved by[CISO and executive risk owner]
Version and date[1.0, date]
Review frequencyAt least [annually], and after a significant incident, failed recovery exercise, or major platform change
Related documents[Business continuity plan; disaster recovery plan; backup standard; data retention schedule; encryption policy; incident response plan]

Implementation guidance

Rollout

  1. Tier first: agree recovery tiers with system owners from the business impact analysis before buying storage. Tiering usually shows that identity infrastructure and SaaS data are missing from backups.
  2. Separate the backup identities: move backup administration off the production directory and rotate backup service credentials before enabling immutability, so an attacker cannot already hold the keys.
  3. Enable immutability in audit first: apply object lock or vault lock in a governance or test mode on a subset, confirm storage growth and restore behaviour, then switch to the enforced mode. Enforced modes generally cannot be reversed during the retention period, so confirm retention values before enforcing.
  4. Prove a clean restore: run a full restore of one Tier 1 service from the isolated copy, timed, with no production credentials. Fix the gaps it reveals before scaling.
  5. Roll back safely: keep existing backup jobs running unchanged until new jobs have passed [two] restore tests.

For platform-specific settings, see the Backup Security and Immutable Backup Guide, the AWS Backup Security Guide and the Azure Backup and Recovery Services Vault Security Guide.

Failure patterns this policy is written to prevent

  • Backups and their key taken together: in the GoTo breach, encrypted backups were stolen along with an encryption key for part of them. Section 4.5 keeps keys apart from the backups they protect.
  • Same credentials for production and backups: attackers who gain domain administrator rights delete backups first. Section 4.5 separates the accounts.
  • Retention by count: repeated corrupted backups push good versions out. Section 4.3 uses time-based retention.
  • Untested restores: a successful job log does not prove a restore works within the RTO. Section 4.6 requires timed restores.
  • SaaS assumed safe: provider resilience does not protect against deletion by a compromised admin or a malicious integration.

Framework mapping

FrameworkReferencePolicy sections
NIST CSF 2.0PR.DS-11 (backups created, protected, maintained and tested), RC.RP-03 (backup integrity verified before restoration)4.2 to 4.6
NIST SP 800-53 Rev. 5CP-9 System Backup, with CP-9(1) Testing for Reliability and Integrity, CP-9(3) Separate Storage for Critical Information, CP-9(5) Transfer to Alternate Storage Site, CP-9(8) Cryptographic Protection; CP-6 Alternate Storage Site; CP-10 System Recovery and Reconstitution4.1 to 4.6
CIS Controls v8.1Control 11 Data RecoveryWhole policy
ISO/IEC 27001:2022Annex A 8.13 Information backupWhole policy

Sources for the default values

Related NHI Mgmt Group resources: Disaster Recovery Plan Template · Backup Security and Immutable Backup Guide · Backup Identity Isolation Guide · Business Continuity and Disaster Recovery Guide · Backup and Cyber Recovery Buyer's Guide

PDF for subscribers

Take this guide with you

Every section, table and checklist in one PDF, with links to the sources and related guides. Subscribe to our weekly NHI & AI Identity Journal and we’ll email you a download link.

Free weekly newsletter

Subscribe to the NHI & AI Identity Journal

The latest on NHI and Agentic AI security – articles, research, breaches, news and events every week.

Bonus 33% off our NHI Course when you subscribe.

NHIMG Editorial Note
Researched and written by the NHI Mgmt Group editorial team using AI tools, under the editorial direction of Lalit Choda. Facts are drawn from the sources listed on this page and checked by our automated quality process. Last updated 27 October 2026. Spotted an error? Use our contact page.
NHI Mgmt Group, the independent authority on Non-Human Identity, IAM and Agentic AI security. nhimg.org

PDF Downloads for Newsletter Subscribers

Get the guide

Enter your details and we’ll email you a download link.

The full guide is free to read online. The PDF is a thank you for subscribing.

Fill in all fields marked * and tick the PDF box to continue.