Browse all topics
Microsoft Defender (Security)

Ransomware preparedness for Microsoft 365

How to harden a Microsoft 365 tenant against ransomware — prevention, detection, response, and recovery.

Ransomware against Microsoft 365 tenants targets cloud content directly — encrypting SharePoint sites, OneDrive accounts, mailboxes, or destroying access via credential theft and admin-account compromise. Even cloud SaaS doesn't make you immune. The good news: Microsoft 365 has strong defences built in, and Microsoft provides specific tooling for recovery. The hard part is having it ready before it happens.

Prevention

Identity hardening

  • MFA enforced for everyone, phishing-resistant (FIDO2 / passkeys) for admins.
  • Conditional Access blocking legacy auth, requiring device compliance.
  • PIM for every privileged role — no standing Global Admins.
  • Identity Protection policies blocking high-risk sign-ins.
  • Break-glass admin accounts stored offline, MFA-exempt but with very strong unique credentials.

Endpoint hardening

  • Defender for Endpoint onboarded everywhere, Tamper Protection on.
  • Attack Surface Reduction rules in block mode for high-confidence rules.
  • Local admin removed for most users (Intune EPM for elevation).
  • Patch SLA kept short — Defender Vulnerability Management drives priority.

Email hardening

  • Defender for Office 365 P2 with Strict preset policies.
  • DMARC at p=reject.
  • Attack Simulation Training running monthly.
  • OAuth consent restrictions — limit user consent to risky scopes.

Detection

  • Defender XDR monitoring with alerting on critical signals.
  • Automatic Attack Disruption enabled.
  • Microsoft Sentinel for cross-product correlation and longer retention.
  • Microsoft Defender for Identity sensors on every domain controller.
  • Unusual activity alerts for mass-download patterns, mass-share-link creation, unusual sign-in countries.

The earlier you detect, the smaller the impact. Microsoft's case data consistently shows that fast disruption changes outcomes dramatically.

Response

When ransomware is suspected or confirmed:

  1. Isolate affected devices via Defender for Endpoint.
  2. Disable affected accounts in Entra ID, revoke sessions.
  3. Stop active sharing links — sometimes mass-share is the attack pattern.
  4. Snapshot SharePoint / OneDrive sites that are still healthy to capture recoverable state.
  5. Engage Microsoft Support — for severe incidents, Microsoft has incident-response support.
  6. Document the timeline as you go.

Recovery

This is where many organisations were unprepared in past incidents:

Native SharePoint and OneDrive recovery

  • Version history — every file in SharePoint and OneDrive has version history. For ransomware that encrypts files, previous versions are typically still healthy.
  • Recycle bin — two-stage recycle bin retains deleted content for 93 days.
  • Restore your OneDrive — built-in feature lets users restore their OneDrive to a point in time (up to 30 days back). The single most useful per-user recovery feature.
  • SharePoint site restore — admin can restore a site to a point in time.

Microsoft 365 Backup

For larger and structured recovery, Microsoft 365 Backup provides:

  • Point-in-time restore at SharePoint / OneDrive / Exchange scale.
  • Hourly snapshots retained per configured policy.
  • Rapid restore measured in items / hour — orders of magnitude faster than typical third-party restore.

For tenants subject to ransomware risk, Microsoft 365 Backup is increasingly part of the baseline.

Third-party SaaS backup

For air-gapped off-platform copies (insurance against a tenant-wide compromise), third-party SaaS backup tools (Veeam, Druva, AvePoint, Acronis, Barracuda, others) provide additional resilience.

The right answer for most organisations is both: Microsoft 365 Backup for fast in-platform recovery, plus a third-party off-platform copy.

Operational readiness

The single most underrated activity: test restoration. Run a quarterly tabletop exercise where you actually restore a SharePoint site, a OneDrive, a mailbox to verify the procedure works. The only backup that matters is one you've recovered from.

Document the runbook. Train the team. Hope you never need it.