Doppel's broader point matters more than any single lure. In its telemetry, the overall amount of phishing did not rise after any of these events. It didn't need to. Credential-phishing kits, fake support numbers and bank lookalike domains are always running. A headline doesn't create more of them. It makes the existing ones look like exactly the message people are waiting for.
The headline parade, October 1 to November 18
According to Doppel, the period produced an unusually dense run of events that gave scammers something to talk about. Doppel lists them as context for its thesis. Apart from the Microsoft 365 campaign below, this article has not independently verified each one, so treat them as Doppel's account.
- Azure multi-region outage (Sept. 30–Oct. 1): Doppel says it lasted about 5 hours 45 minutes. Paused gateway updates resumed all at once and overloaded the networking control plane.
- Denmark population-register breach (Oct. 5): About 8.8 million people were affected. Access came through a private company's legitimate lookup rights.
- IDCF Cloud ransomware (Oct. 7): An attack on IDC Frontier, a SoftBank-group cloud platform in Japan, disrupted 495 companies and local governments.
- Supermarket refrigeration sabotage (Oct. 11–12): Attackers forced controllers at French supermarkets into defrost cycles, according to Doppel.
- ConnectWise Cloud Backup credential (Oct. 15–16): The credential held standing consent into customers' Microsoft 365 tenants. Doppel says it was used to attempt sign-ins to about 1,000 tenants and succeeded in more than 300. In at least two tenants, a related app gave itself the global administrator role.
- Binance theft (Oct. 29): About $6.3 billion was taken from hot wallets, which Doppel calls the largest cryptocurrency theft on record. Doppel traces it to a tampered developer installer and a fake job interview months earlier.
- FBI FLASH on ARTEX (Nov. 13): ARTEX is an open-source autonomous penetration-testing framework that criminals have reportedly used since August. Doppel says some intrusions reached privileged access in under three hours.
- Microsoft 365 outage (Nov. 12): Exchange Online, Outlook and Teams were down for most of a business day.
- Nationwide FAA ground stop (Nov. 13): Doppel says operators canceled about 12,462 flights.
- Optimizely script compromise (Nov. 16): An injected loader ran on the sites of 178 organizations for about 19 hours. It skimmed cards and swapped cryptocurrency wallet addresses.
- US bank software bug (Nov. 16): Transactions at one of the largest US banks slowed for most of a day. The bank says the cause was not cyber-related.
- Critical Java flaw and NSANet maintenance (Nov. 17–18): Doppel notes that NSA's classified network degraded during maintenance on the same day as the Java disclosure.
Doppel says each event followed the same pattern. A domain named after the event appeared within hours. A page asking for personal or financial details appeared within a day. Everything else was the standing supply of lures, now with a believable backstory.
The Microsoft 365 outage: a case study in speed
This is the part Microsoft shops should read closely.
Microsoft's own findings. Microsoft confirms that on November 13, 2026, Microsoft Threat Intelligence observed a coordinated, high-volume phishing campaign targeting organizations across multiple sectors worldwide. The emails used the subject pattern "Outage – <random token>" and display names built from a three-character, organization-style prefix followed by "Outage Notification" or "Service Health." The bait claimed that mail had been held during the outage and would be deleted within 24 hours. The buttons said "RESTORE MESSAGES" or "TAKE ACTION," and the messages mentioned Exchange Online, Outlook, Teams, SharePoint and Copilot.
The attackers also used a real incident number. Cybersecurity News reported Microsoft's November 14 warning on X, which said the messages impersonate Microsoft 365 service health notifications and reference the legitimate Exchange Online incident EX1550543. That publication also stressed that the phishing campaign should not be confused with its cause. Microsoft's outage updates said there was no indication that malicious activity had caused the underlying service failure.
The dates are easy to mix up:
| Milestone | Date |
|---|---|
| Outage begins (first advisory 16:17 UTC, per Doppel) | Nov. 12 |
| Microsoft observes the phishing campaign | Nov. 13 |
| Microsoft Threat Intelligence posts public warning | Nov. 14 (06:32 UTC, per one report) |
| Microsoft Security blog analysis published | Nov. 16 |
Microsoft was careful about what it could prove. It assessed the activity as likely credential phishing. It said it had no evidence of credential capture, malware delivery or post-compromise activity, and it did not attribute the campaign to a named actor. The sample set contained only links: no QR codes or malicious attachments. The named sender domains were moments-notice[.]com, planetmaison[.]com and newbooker[.]com. Tracking links in multiple samples led to reallifeaccounting[.]com, and some samples went to outagefixed[.]sbs and outagerestored[.]sbs.
Independent corroboration. Mimecast's own visibility matches Microsoft's account. The email-security vendor wrote that within hours of the Microsoft 365 service disruption on November 12, 2026, we began seeing phishing emails that use the outage as a lure. Mimecast also saw extra variants, with display names such as "Service Health", "Mail-Restore-Auth" and "Health Status" and subjects that begin "Outage -" and "Restore –", "Strengthen-Security". Microsoft's indicator list includes the "Strengthen – Security" and "Restore — 4 messages" subject patterns too. Two vendors seeing the same thing makes the campaign hard to dismiss as a one-off.
Doppel's timeline from the domain side:
- 6 hours 20 minutes after Microsoft's first advisory:
microsoftoutage[.]comwas registered. It had live content within 2.5 hours, but Doppel describes it as a revenue-loss counter with no forms. It was not verified as malicious. - 10 hours 24 minutes:
outlookoutage[.]comappeared, hosting a "community forum" with its own account sign-up. - 23 hours:
winrefrequests[.]comwent up with a "Microsoft Refund Request Form." It asked for full name, date of birth, Microsoft account, alternate email, phone, billing address, payment method, bank name and refund amount, and sent the answers to a relay on a second domain. Doppel found nothing connecting it to the service-health email campaign. In its words, two crews worked one headline with two different asks.
Doppel also reports a second email campaign on November 17 with the subject "disconnected after the outage." It linked to a sign-in clone that walks the user through entering a verification code. Doppel says this fits either a device-join attempt or plain credential theft. The same "outage" banner appears on the sign-in page whatever the lure, which suggests a kit built to be reused.
Summary
A real outage plus a real incident number made a very believable phish. The campaign started within a day, and two security vendors independently saw it in customer mail.
The finding that matters: no spike, higher hit rate
In public scan data, Doppel counted pages mentioning the outage rising from about two a day to 27 in the first 24 hours. Sixteen of those were news coverage, and none collected credentials. Meanwhile, Microsoft sign-in impersonation pages on third-party hosting ran at about 40 over the same 26 hours. That was below the previous week's daily range of 43 to 163.
So nobody needed to build new phishing pages for the outage. The existing ones were already offering what more than 400 million Microsoft 365 users wanted as service came back: a way back into their mail.
Doppel saw the same pattern elsewhere:
- Bank slowdown: New lookalike hosts using the names of large US banks stayed within the previous month's range of 2–46 a day. None of the 32 was confirmed as live impersonation.
- Binance: Lookalike hosts held steady at one to twelve a day after the theft. With low confidence, Doppel expects recovery scams to show up first on support channels and social platforms rather than on lookalike domains.
- Ground stop: Fake airline support-number pages had been posted in bulk the day before the ground stop. Scammers hadn't predicted the event. They keep those pages running all the time, so they were ready when it happened.
This is a SOC problem. If your detection waits for a spike in phishing volume, an event like this won't trigger it. The volume stays the same; what changes is how many people click.
The usual caveat applies. Doppel sells brand-protection services, and its telemetry covers its own customers, strongest in North America. The "no measurable rise" finding is Doppel's observation, not a measurement of the whole industry. Still, it fits how commodity phishing works. Phishing-as-a-service kits run all the time, and headlines just give them better cover.
What Microsoft 365 admins should do now
Microsoft's guidance in its November 16 analysis is practical:
- Hunt and purge. Search mail telemetry for the sender domains, subject and display-name patterns, lure wording and destination domains. Quarantine or remove matches according to your response procedures. Microsoft provides Defender XDR advanced-hunting queries against
EmailEvents,EmailUrlInfoandUrlClickEvents. They have no preset time range, so set one in the hunting interface. The window from November 12 onward is the obvious place to start. - Find the clickers. Use click telemetry, proxy logs and endpoint or browser evidence to see who reached the destination and whether anyone may have entered credentials.
- Remediate likely victims. Reset passwords, revoke sessions and refresh tokens, and check MFA registrations. Review sign-ins, OAuth consents, inbox and forwarding rules, delegate permissions, and unusual Exchange Online, SharePoint, OneDrive or Teams access.
- Block carefully. Block or closely monitor
reallifeaccounting[.]comand the related paths. Don't broadly block shared delivery infrastructure: the tracking hosts include SendGrid subdomains, and blocking them could catch legitimate mail. - Teach the verification rule. Microsoft is explicit: legitimate Microsoft 365 service-health communications do not direct them to non-Microsoft domains to recover held mail. Validate service issues through the Microsoft 365 admin center rather than through unsolicited email links.
- Harden for the next event. Use phishing-resistant MFA such as passkeys or security keys. Review Safe Links and Safe Attachments, enable Zero-hour Auto Purge, use attack simulation training, and make sure users report suspicious mail through Outlook.
Doppel's advice applies beyond Microsoft 365. Publish your single status page, support channel and refund route before the first lure arrives. Tell help desks that "sign back in" and "your mail will be deleted" are the phish to expect after any outage. Never take a support number from a search result. Inventory standing third-party consents in your tenant: the ConnectWise incident is a reminder that an OAuth grant from years ago can be the way in.
Summary
Treat every outage as a phishing alert. Microsoft has published indicators and hunting queries. The bigger defense is having users check status in the admin center or the app itself, never through a link in an email.
The bottom line
The deadlines aren't slowing down either. Black Friday is November 27 and Cyber Monday is November 30. Doppel notes that holiday-deal lures, ground-stop refund offers and Microsoft 365 sign-in prompts are now all competing for the same inbox. The lesson from these seven weeks is that attackers didn't need to build anything new. They waited for the next headline and used lures they already had.
So the question for admins isn't whether a spike will show up on the next outage. It's whether your users already know where the real status page is.
References
- Never-Ending November - Security Boulevard Security Boulevard · 2026-11-19T19:49:04+00:00
- Attackers Exploit the Microsoft 365 Outage with Fake “Held Messages” Lures mimecast.com
- High volume phishing campaign uses Microsoft 365 outage lures to target multiple industry sectors microsoft.com