What happened: Microsoft observed a high-volume wave of fake service notices on November 13, 2026, as people dealt with a real Microsoft 365 disruption.
What to do: Do not use an unexpected message to restore mail or sign in. Open your usual Microsoft app or contact your IT team directly. If you entered a password or approved a sign-in, act promptly.
The Microsoft 365 Outage Was Real. The Email Warning Was the Trap.
After Microsoft 365 went down, attackers sent messages claiming delayed emails would disappear unless recipients acted. Microsoft says the campaign likely sought credentials, but has not confirmed that anyone submitted them.

Jump to a section
Imagine your inbox has been unreliable. A message arrives saying several emails are stuck and will be deleted unless you restore them. It sounds like the update you have been waiting for.
That is the opening the attackers wanted. On November 16, Microsoft Security reported that it had observed a coordinated, high-volume phishing campaign on November 13. The emails posed as Microsoft 365 service-health notices and warned that held messages would be deleted within 24 hours.
Microsoft assessed the campaign was probably designed to capture credentials. Its published evidence did not confirm that anyone entered credentials, that malware was delivered or that accounts were compromised. The company did not name a responsible group or provide a victim count.
The service disruption itself was real. Our coverage of the Microsoft 365 outage explains how problems affecting email and connected services spread across organizations. A phishing campaign exploiting the disruption is separate from its cause: Microsoft said it had no indication that malicious activity caused the outage.
The warning that made the outage useful to attackers
Microsoft says the messages impersonated service-health notifications, used randomized outage-themed subject lines and directed recipients through tracking links to a non-Microsoft destination. The messages claimed email had been held during the disruption and pressed recipients to restore it before a deadline.
Separately, security company Field Effect examined an outage-themed email detected on November 13. In that example, the message claimed four emails were being held, referenced the real Exchange Online incident number EX1550543 and contained a button labeled “RESTORE MESSAGES.” Field Effect said the button led to a page intended to collect credentials. Its description of that message does not establish how many people submitted information.
The detail that makes this convincing is also the detail that makes it dangerous: the service really had been disrupted. A real incident number can be copied into a fake notice. A familiar logo, an outage you experienced and an urgent deadline do not establish that the sender is trustworthy.
Another security company, Mimecast, reported seeing outage-themed emails within hours of the November 12 disruption. Its examples included both a held-messages notice and a separate security-check prompt. The different versions matter: blocking one subject line or button is not enough to identify every suspicious message.
Microsoft, Field Effect and Mimecast describe related outage-themed lures. Their publications do not establish that every outage-themed email or support call came from the same operator, or that every version had the same destination.
The second problem after an outage
When a service fails, people look for instructions. They may be waiting for a missing invoice, a customer response or a meeting invitation. An attacker does not have to invent that concern. The message only needs to make an unsafe action look like the next step in getting back to work.
That is the overlooked risk: an outage changes what people expect to see. A notice that would look strange on an ordinary day can feel plausible when the inbox has just stopped working. The same pressure can travel by text or phone, not just email. Field Effect also describes reports of outage-themed support impersonation and payment-redirection requests. Those are warnings about possible variants, not proof that every such contact belongs to Microsoft's observed email campaign.
Our earlier reporting on fake AI event invitations targeting work accounts shows another version of this trust problem. There, a useful-looking event was the lure; here, it was a real disruption. The campaigns are separate, and the reporting does not establish a shared operator.
What to do with a message about missing mail
- Do not use the message's button or phone number. Open the Microsoft app or website you normally use, or contact your employer's IT team using a known channel. Administrators can check Microsoft service health and their Microsoft 365 admin center.
- Check the request, not just the branding. Be wary of a short deletion deadline, a demand to sign in again or an unexpected offer to fix your account. A real incident number is not proof that a particular message is genuine.
- Report the message. For a work or school account, use your organization's phishing-reporting process or contact IT directly. Do not forward it to colleagues as a warning unless your security team asks you to.
- Verify payment changes separately. If a supplier or colleague says the outage requires payment to a new account, call a number you already trust before sending money.
Some organizations do send genuine quarantine notifications about messages held by security filters. Those are different from email delayed by an outage, and the actions available depend on the organization's quarantine policy. If you are unsure whether a message is actually being held, ask IT or use your organization's approved portal directly. Do not release an unfamiliar message simply because it appears in quarantine.
If you already clicked, signed in or approved a request
Received it but did not interact? Report it through a trusted channel and delete it when your organization's process permits. Receiving a message does not by itself mean your account was compromised.
Clicked but did not enter information or download anything? Close the page and tell your IT team if it is a work account. A click alone does not prove a takeover, but your security team may want to check the destination and any device warnings.
Entered a password, code or approved a sign-in you did not initiate? Contact your work or school IT team immediately using a known channel. For a personal Microsoft account, use a trusted device and Microsoft's compromised-account guidance to secure access. Change an exposed password through the official service, not through the message, and change it anywhere else you reused it.
Installed software or allowed remote access? End the session if it is safe to do so, stop using that device for sensitive tasks and contact IT or a trusted technician from another device.
Supplied payment details or sent money? Contact your bank or payment provider immediately through a trusted number. Tell them what you shared and ask what can be stopped or disputed.
A password change may not be the whole cleanup
Email is more than a mailbox. It can contain password resets, financial conversations and links to other services. If an attacker gained access, a new password is one step, but the organization also needs to check for access that may have been left behind.
Microsoft's guidance for compromised Microsoft 365 accounts tells administrators to revoke sessions and review authentication methods, app access, forwarding and inbox rules. Those checks matter because an attacker who obtained access may have changed settings to keep receiving messages or regain entry. That is a general risk, not something Microsoft confirmed happened in this campaign.
Our Outlook and Microsoft 365 security guide walks through these checks. If you cannot sign in, use the appropriate personal or work-account steps in our Microsoft account recovery guide.
The outage made a restore notice feel timely. The safer question is not whether the disruption was real, but whether you independently verified the request that followed it.