Have you been compromised? Call our active incident line at 1-650-300-7557 or see how we help.

,

Microsoft 365 outage security checklist: during and after an incident

A Microsoft 365 outage is an availability incident, not proof that your tenant was attacked. It still creates security work: users look for shortcuts, criminals imitate recovery messages, and some Microsoft security portals or controls may be slow or unavailable. The safe response is to preserve identity controls, use only approved continuity methods and review the incident window after recovery.

Incident status rechecked November 15, 2026: Microsoft lists EX1550543 and MO1550821 as service restored, with an incident end time of November 13 at 8:10 AM Pacific. Its final status describes a technical issue introduced during an earlier network-capacity upgrade and surfaced by a later network change—not cyber activity or AI agents. A full post-incident report is pending. Use the checklist for recovery review and the next outage; check Health > Service health for any current tenant issue.

First decide: outage symptoms or signs of compromise?

Microsoft’s Exchange incident covers Outlook on the web, desktop MAPI, Exchange ActiveSync and REST-based mailbox access, with possible delivery delays. The wider incident lists feature-specific effects across other Microsoft 365 services. Widespread connection failures that match the advisory are different from evidence that an attacker accessed one of your accounts. Do not force every observation into one explanation.

ObservationAppropriate first response
Many users cannot connect through more than one normal Outlook methodCheck EX1550543 and MO1550821, record the time and affected workflows, and avoid tenant changes as an outage fix.
A user receives an unexpected ‘restore your mailbox’ link or MFA promptDo not use the link or approve the prompt. Report it through the normal security route and navigate only through an established Microsoft sign-in path.
An unfamiliar sign-in, device, MFA method, forwarding rule or app consent appearsTreat it as a potential security incident. Preserve the evidence and follow your account-compromise procedure rather than dismissing it as an outage symptom.
Safe Links, a Defender portal or a Purview page is unavailableRecord the gap and use compensating caution. Do not disable other controls or assume that no alerts means no activity.

During the outage: the first eight checks

  1. Check Service Health. An administrator should review the two incidents under Health > Service health. Use Microsoft’s public status route or Admin app if the admin center is unavailable. Record the update timestamp.
  2. Assign status and security owners. One person coordinates service updates; another handles suspicious messages and account evidence. They can be the same person in a small team, but the responsibilities stay explicit.
  3. Send one approved staff notice. Explain that users should not reinstall apps, remove accounts, disable MFA, forward work to personal email or install recovery tools. Name the approved out-of-band channel and next update time.
  4. Keep help-desk verification intact. A busy queue is not a reason to skip identity proofing for a password reset, new MFA method or remote-access installation.
  5. Inventory critical work. Record delayed messages, files, scheduled meetings, approvals, security alerts and external integrations. Give every urgent item an owner and a safe next action.
  6. Use only existing offline access on managed devices. Previously cached email, drafts and synced files may remain available. Do not create new local copies on home or unmanaged devices to get around the outage.
  7. Test the exact feature, not the brand. Microsoft said direct SharePoint file access and core Teams messaging/meetings remained largely functional at the checked update, while Exchange-dependent features could fail. A successful test in your tenant is not a guarantee for every user or feature.
  8. Write down every exception. Record any temporary process, who approved it, the information involved and how it will be reversed or reconciled. The safest target is no new security exceptions at all.

Do not turn a Microsoft outage into an identity incident

MFA, Conditional Access, mail-flow rules, data-loss prevention and sensitivity labels do not repair Microsoft infrastructure, and disabling them creates risk that can outlive the outage. The same applies to automatic external forwarding, personal cloud accounts and unapproved remote-access tools.

CISA’s Scattered Spider advisory describes criminals impersonating staff and help desks to obtain password resets, MFA transfers and one-time codes. Use that general lesson to maintain caller verification and phishing-resistant MFA. It is not evidence that Scattered Spider or any other attacker caused this incident.

Emergency access is preparation, not an outage shortcut

Organizations should maintain governed emergency access before they need it. Microsoft recommends more than one emergency account, phishing-resistant credentials, secure storage, monitoring and regular validation, with deliberate exclusions from policies that could block emergency sign-in. That is very different from switching off Conditional Access for everyone. Review Microsoft’s emergency-access account guidance during planning, and use those accounts only for their documented emergency purpose.

Treat reduced security visibility as a risk to manage

At the checked update, Microsoft said Safe Links might be unable to verify URLs. It also listed possible effects on Defender portal pages and several Purview functions, Intune and Exchange-backed Microsoft Graph requests. A quiet console or delayed alert therefore needs an explicit review, not an assumption that nothing happened.

  • Warn users to avoid unexpected recovery links, attachments and MFA approvals even when a message appears to come from IT.
  • Keep independent endpoint, identity and network alerting paths in view where your service includes them; note any telemetry that is also delayed.
  • Record which security functions were unavailable and when, so the later review covers the whole gap.
  • Do not remove a sensitivity label or relax DLP merely to complete an export. Microsoft’s own workaround includes the condition that organizational data-protection policies must allow it; deferring the task is often the safer choice.

Secure workarounds that can help

NeedLower-risk option to testBoundary
Read or draft emailPreviously synchronized content in classic or new Outlook on a managed device; save drafts or Outbox items for later.New offline settings cannot retrieve mail that is still only on the server. Review queued items before sending or resending.
Work on filesAn existing local or synced copy on an approved workstation, with one editor and a version log.Sync is not independent backup. Do not create uncontrolled duplicates or move data to personal storage.
CommunicateThe organization’s pre-approved phone or out-of-band process; test core Teams functions if they work in the tenant.Do not build a new consumer-chat process around the incident or put sensitive details into an unapproved channel.
AdministrationService Health through the admin center, public status route or Admin app; documented emergency access only for its intended scenario.Do not make broad tenant configuration changes as a workaround for a server-side incident.

Microsoft describes new Outlook offline access and classic Cached Exchange Mode separately. Confirm which client and existing local data a user has rather than issuing one instruction to everybody.

After service returns: the ten-point tenant review

  1. Verify the tenant incident status and real workflows. Test sign-in, mailbox access, sending, receiving, files, calendars and important integrations rather than relying only on a green icon.
  2. Trace delayed messages. Use Exchange admin center > mail flow > message trace for specific messages. Microsoft notes that trace data can take time to appear.
  3. Review queued and duplicate messages. Check Outboxes and repeated manual sends before creating another copy.
  4. Reconcile offline files. Resolve versions, permissions and linked files with the people who worked on them.
  5. Review identity activity. Check unfamiliar sign-ins, repeated failures followed by success, new devices, MFA registrations and emergency-account use.
  6. Review mailbox persistence. Look for new inbox rules, forwarding addresses, delegates and app consents created during or just after the incident.
  7. Review security and compliance gaps. Check late Safe Links, Defender, Purview and Intune alerts and confirm expected policy enforcement where your licenses provide those functions.
  8. Validate independent monitoring. Confirm that external endpoint, identity, network or other telemetry paths are current and that delayed data is accounted for.
  9. Close temporary procedures. Reverse every exception and reconcile manual approvals, payment instructions, client messages and support tickets.
  10. Document outcomes and wait for the final report. Record actual impact and corrective actions. Update the incident record when Microsoft publishes its post-incident review rather than treating an early working explanation as final root cause.

Prepare the next response before the next outage

A useful outage playbook names the incident owner, approved communication route, offline-access rules, identity-proofing requirements, evidence checklist and recovery order. Exercise it with realistic conditions: a Microsoft portal that will not load, a user asking for an MFA exception, a delayed security alert and a project file edited in two places.

Start with our Microsoft 365 security baseline for identity, mailbox, sharing, logging and backup fundamentals. For broader business-continuity planning, coordinate the technical checks with the people who own each critical workflow; a green backup job is not the same as a successful restore or a workable downtime process.

For engineering, architecture, construction and manufacturing teams, see the CloudWorks project-continuity guide for local files, approved communication, supplier verification and recovery reconciliation. It complements—not replaces—the tenant security review.

Review your outage security plan with ACS Cyber

ACS Cyber can help you map status ownership, identity verification, independent monitoring, incident-window evidence and recovery actions. Our managed detection and response service provides 24/7 managed detection and response on the endpoints we protect, with ACS engineers handling escalation, containment decisions with you, recovery and reporting. Coverage can include Microsoft 365 and other identity, network and cloud activity, while our vCISO advisory helps turn lessons into an owned security roadmap. Exact coverage and response responsibilities are agreed with each client.

Bring your current Microsoft 365 configuration, continuity procedure and last incident or restore-test record to a free 30-minute consultation. We will identify practical next steps without promising that every future outage or attack will have the same outcome. Talk to a Cybersecurity Expert.

Suspect an actual account or business compromise rather than only a cloud outage? Active incident? Call 1-650-300-7557 now and say it is an active incident, and we will bring in our security team. After-hours emergency support is available, and you do not need to be an existing client.

Frequently asked questions

Should we disable MFA or Conditional Access until Microsoft 365 is stable?

No. Those controls do not fix a Microsoft infrastructure incident, and weakening them creates an opening for account compromise. Preserve normal identity verification and prepare monitored emergency access in advance for the specific lockout scenarios it is meant to address.

Is an Outlook outage a sign that our tenant was breached?

Not by itself. Microsoft’s advisories describe an availability incident. Investigate unfamiliar sign-ins, MFA changes, forwarding rules, app consent or unexpected recovery messages separately on their own evidence.

Can we switch from desktop Outlook to web or mobile?

You can test your normal clients, but Microsoft listed Outlook on the web, desktop MAPI, Exchange ActiveSync and REST as potentially affected. Switching clients is not a reliable fix. Existing offline content may still be useful on a managed device.

What is the first security check after recovery?

Confirm the actual workflows and delayed mail first, then review the incident window for identity changes, mailbox rules, app consent and late security or compliance alerts. Record what you verified and what remains unresolved.

Want a second opinion on your security?

Talk with an ACS Cyber specialist about your environment, your obligations and the next practical step. Dealing with an active incident? Call our incident line at 1-650-300-7557 (after-hours emergency support is available).

Keep reading

More practical guidance

Your privacy choices

We use a small set of tools to understand how this site is used and to follow up on consultation requests. You can turn optional analytics and marketing cookies off for this browser at any time.

We also keep simple counts of scheduler opens, bookings and calls by page and button. When you book, open our booking calendar on calendly.com, or tap our phone number, we also save which pages of this site you read and for how long, with a random browser ID, so our team can follow up well. Turning analytics off here (or sending GPC) stops this. We honor the Global Privacy Control (GPC) signal. Details are in our Privacy Policy.