Microsoft Blames July Software Defect for 365 Outage

Microsoft's preliminary post-incident report, dated 15 November 2026, says the Microsoft 365 outage of 12 and 13 November came from a latent software defect in a fix deployed to its networking gateway servers in July. The defect stayed hidden for more than three months. It was triggered when Teams traffic moved onto a gateway shared with Exchange Online, while about 14% of that gateway's capacity was out of service. Outlook, Teams, SharePoint and Copilot were disrupted in many countries. Microsoft says no cyberattack was involved and that customers do not need to change their tenant settings. The final report is due on 20 November 2026. On 16 November, a proposed Canada-wide class action was filed in the Ontario Superior Court of Justice. The allegations have not been tested in court.
In short
- Microsoft's preliminary report blames a July software fix defect, a Teams traffic shift and about 14% lost gateway capacity.
- The outage lasted almost 24 hours and affected Outlook, Teams, SharePoint and Copilot in many countries.
- Automated routines that rely on Exchange Online or Teams need a written fallback and a visible failure log.
- A proposed Canadian class action was filed on 16 November 2026; it has not been certified or tested in court.
- Microsoft's final post-incident report is due on 20 November 2026.
What happened
According to Microsoft's preliminary post-incident report, dated 15 November 2026, the Microsoft 365 outage began on the afternoon of 12 November UTC and ended at 16:10 UTC on 13 November, lasting almost 24 hours. The report traces the failure to a fix that Microsoft built for its own networking gateway servers and began deploying in July 2026. Under certain loads, that fix caused an internal Windows process to consume the servers' processor and memory. Microsoft says the problem is limited to that gateway infrastructure and does not affect other Windows systems.
Two further conditions set it off. Roughly 15 hours before the outage, Teams traffic failed over from the alternative gateway to the primary gateway that also supports Exchange Online. Separately, an unrelated issue had already taken about 14% of the gateway's server capacity out of service. Microsoft executive Charles Lamanna described these three factors in a public post on 16 November. At the worst point, more than 80% of the affected gateway servers stopped responding, according to the report.
Downdetector, the outage tracker, recorded reports from users in 74 countries, according to ZDNET. Email was the hardest hit. Teams chats and meetings worked for most of the outage, according to Microsoft's own account.
Why it matters for businesses that automate work
Most automated routines depend on the same mail, calendar and file services that stopped working on 12 November. Follow-up emails, invoice sends, document requests and approval chains all route through Exchange Online, SharePoint or Teams. When those services fail, automated steps that expect a reply can stall, queue or fail without anyone noticing at once.
The incident also shows that a fault can sit unnoticed for months. Microsoft says the defect ran in production for more than three months before the extra load and lower capacity exposed it. A business that tests its automations only during normal operation may not learn how they behave under stress.
What changes for a business that automates work, in practice
For most businesses, the first change is a list. Write down which routine tasks depend on Microsoft 365 and which of them would cause a customer or a supplier to wait if they stopped for a day. Payment reminders, purchase order emails and HR notices usually sit near the top.
The second change is a fallback. A simple table is enough for a small firm:
| Routine task | What happens if mail or Teams is down | Fallback to agree in advance |
|---|---|---|
| Payment reminders | Reminders are not sent | Hold a queue and send once service returns, or call the top ten accounts |
| Document requests | Staff cannot collect files | Use a shared portal or a messaging app the team already has |
| Approvals | Approval emails stall | Name a second approver who can act by phone |
| Customer notices | Updates are delayed | Post a status message on the website or a second channel |
The third change is to check how each automation reports failure. An automation that stops quietly is more dangerous than one that stops loudly. Someone should review the failure log each morning until the vendor's final report is read.
Automation cannot fix a vendor outage. It can hold work, route it to a person, and show what was not sent. Whether any of this changes a business's exposure depends on its own volumes and contracts, which a free automation audit can measure. Businesses should also read their Microsoft 365 service agreement to see what it offers during an outage, since service terms differ between plans.
The Canadian class action
On 16 November 2026, Diamond & Diamond Lawyers filed a proposed Canada-wide class action in the Ontario Superior Court of Justice against Microsoft Canada Inc. and Microsoft Corporation, according to a news release issued through CNW. The release says potential economic losses could exceed $100 million. It also says the proposed class has not yet been certified by the court. A separate firm, Consumer Law Group, said on 16 November that it is investigating a class action. The allegations have not been tested in court, and Microsoft has not responded to them in the sources reviewed.
What to watch next
Microsoft says its final post-incident report is due on 20 November 2026 and will include a list of repair items. Microsoft has also said that it increased gateway capacity and added safeguards against redirected traffic causing similar failures. Business owners should read the final report in the Microsoft 365 admin center, where administrators can find the incident under identifiers EX1550543 and MO1550821. The court process in Ontario will take longer than the news cycle. Any certification decision would be a separate step.
Sources
How AiStaffo would automate this
AiStaffo would start by mapping which routine tasks depend on Outlook, Teams and SharePoint, such as payment reminders, document requests and approvals. The connected AI workers would then run those tasks as normal, and when mail or chat stops responding, they would hold the pending messages in a queue, log each one as not sent, and alert a named person. The business owner still decides which customers get a phone call, which approver steps in and whether to change suppliers. Book a free automation audit.
Questions people ask
What caused the Microsoft 365 outage on 12 and 13 November 2026?
Was the Microsoft 365 outage a cyberattack?
What is the proposed Canadian class action against Microsoft?
Can a business get compensation from Microsoft for the outage?
What should a business do now to prepare for the next cloud outage?
Book a free automation audit
Thirty minutes. We look at one process you run every week and tell you exactly what an AI worker would take off your desk, and what it would not.

































