Microsoft 365 is degraded worldwide, with Exchange Online the worst-hit service, after a large number of Microsoft servers became unexpectedly unavailable and pushed abnormal load onto the remaining infrastructure. Microsoft’s own service health page listed the condition as service degradation for both Microsoft 365 for business and its consumer products at 3:07am UTC on 13 November 2026 — 11:07am in Malaysia.
The disruption began on the afternoon of 12 November on the United States east coast and ran for more than five hours before Microsoft reported a partial recovery. It reached beyond Microsoft’s own products: OpenAI said it was investigating a fault with its Microsoft connectors, and Anthropic said an upstream supplier problem had affected its Microsoft 365 connector.
Last updated: 13 November 2026, 11:45am MYT. Microsoft’s public status page was still showing service degradation at that time; the company had not published a root cause.
Microsoft 365 down: which services Malaysian SMEs lost first
Microsoft’s status page carried a single line for the business suite — “Users may be unable to connect to multiple Microsoft 365 services” — and the same degradation notice for consumer products. The detail sits behind the admin centre, which is the problem: the page itself notes that administrators may be unable to sign in to read it.
The service list Microsoft confirmed as affected, per its own status updates reported by ABC News, covers Exchange Online, Microsoft Teams, SharePoint Online, Copilot 365, Defender, Planner, To Do, Intune, Purview and Power BI. Exchange Online was named first in Microsoft’s public statements and dominated user reports throughout the incident.
Source: Microsoft Service Health Status, status.cloud.microsoft · page last updated 13 Nov 2026, 03:07:45 UTC · captured 13 Nov 2026, 11:36 MYT
What Microsoft says went wrong
Microsoft ruled out a traffic surge. In its words: “We’ve determined that the elevated load isn’t driven by an overall increase in traffic, but by a large number of servers being unexpectedly unavailable, which is shifting load onto the remaining infrastructure.”
The company said it was investigating why the servers dropped out while restarting and validating each one before returning it to service. It also paused related deployment changes to stop the fault spreading, and began shifting traffic onto alternate routes.
The first public signal came from Microsoft’s own account at 1:54am Malaysian time, which named the mailbox problem and a reference number: “We’re investigating reports of users experiencing issues accessing Exchange Online mailboxes through multiple connection methods in the Americas. More details can be found in the Microsoft 365 admin center under EX1550543.”
Source: @MSFT365Status (X, verified Microsoft account), post of 13 Nov 2026 1:54am MYT · captured 13 Nov 2026 from x.com/MSFT365Status/status/2120938282204688468
An earlier Teams fault in North America, tracked under TM1550488, had already been declared resolved through Microsoft’s automated recovery features before the mailbox problem surfaced.
How large the outage became
Downdetector, which aggregates user-submitted outage reports rather than measuring service availability directly, logged more than 120,000 Outlook problem reports at the peak, ABC News reported, alongside more than 45,000 against Microsoft 365. Reuters put the Outlook count at about 97,000 by 2:07pm US eastern, with more than 38,000 reports against Microsoft 365 and about 4,500 against Teams.
By 5:47pm eastern the Microsoft 365 count had fallen to about 13,000, and Teams reports to 763. That decline tracks Microsoft’s own language of “limited recovery” rather than a clean restoration, and the company said servers returned to service had so far remained stable.
The blast radius included the AI layer sitting on top of the same tenant. OpenAI said new connections through its Microsoft connectors might be broken and that tools were not working as expected; Anthropic said it had identified an upstream provider issue affecting its Microsoft 365 connector. Both are consequences of the same failure, not separate incidents.
What Malaysian SMEs should do while it clears
If your team is locked out this morning, the first question is whether the fault is at your end or Microsoft’s. Check the tenant’s service health in the admin centre and the public status page before rebooting routers or resetting passwords — a local reset during a platform outage adds a second problem to the first.
For mail, the web client and the mobile app often keep working when a desktop client will not, because they use different connection paths. Outbound mail sent during the window is normally queued rather than lost, so avoid re-sending the same message repeatedly; duplicates in a customer’s inbox are harder to explain than a short delay.
Two checks are worth doing today rather than during the next outage. Map which of your business processes stop when a single cloud tenant is unavailable — approvals, invoicing, delivery scheduling, customer replies — and write down the fallback for each. Then confirm whether your accounting or e-invoicing stack depends on the same identity and mail tenant, because that determines whether an outage of this kind is an inconvenience or a deadline you miss.
Malaysian operations were not named as an origin of the fault. Microsoft’s first advisory scoped the mailbox issue to the Americas, then widened it: the company confirmed the problem “may affect Microsoft 365 service connectivity for users in any region”.
Dr Henry Tye, founder and CEO of BigDomain, spoke with Kalvin Chua, Managing Director of EvoPoint — a Malaysian Microsoft Solutions Partner holding designations across Modern Work, Business Applications, Data and AI, and Infrastructure — about this week’s Microsoft 365 outage, which took Outlook, Teams, SharePoint, OneDrive and Copilot offline across 74 countries and was still showing as service degradation on Microsoft’s own status page on Friday afternoon. The advice he gave applies to every Microsoft customer in the country: when a problem hits, go to your incumbent support partner first — that is the escalation path you are already paying for — and only if you cannot reach them should you go to EvoPoint directly at [email protected] or evo-point.com (general enquiries: [email protected]) and see whether they can put a working solution in front of you.
The wider lesson is the harder one. A business today cannot run without email, cannot run without network, cannot run without cloud, and increasingly cannot run without AI — which means disaster recovery is no longer an IT line item, it is a business continuity decision. Ask the questions now, while everything is still up: if your mail died for a full working day, who do you call, and can you still quote, invoice and answer a customer? Do you have a second way to reach your team that does not sit on the same vendor, a fallback DNS or an alternate mailbox, and a backup you have actually restored from — not just one you pay for? Know your RTO and your RPO in hours, not in adjectives. One overloaded health check took out Outlook, Teams, SharePoint and Copilot for most of a day; the businesses that barely felt it were the ones that had already answered those questions.
What is still unconfirmed
Microsoft has not published a root cause, only a mechanism — servers becoming unavailable under load. The reference numbers for the two incidents, EX1550543 and TM1550488, remain the only identifiers in the public record.
No data-loss claim has been made by Microsoft or by any of the affected AI providers, and none should be inferred from a mailbox that failed to sync. Recovery statements at the time of writing described returning servers to service and shifting traffic, not a completed restoration.
Update, 3:45pm Malaysian time: Microsoft’s public status page still reads service degradation for Microsoft 365 Business and Enterprise, four hours after this article was first published. A later advisory narrowed the mechanism: a core process on the affected servers became overloaded, the servers stopped responding, and capacity was lost. It also widened the service list beyond the first one, adding Intune, Purview and Power BI to Exchange Online, Teams, SharePoint Online, Copilot, Defender, Planner and To Do. Microsoft has since said it monitored a change deployed to increase capacity, confirmed that the change did not take effect as expected, and is working on alternatives for what it calls a persistent traffic-allocation imbalance. No root cause and no post-incident report have been published. Downdetector’s peak count was later put at about 138,000 Outlook reports, and the fault was tracked across 74 countries.
Sources
- Microsoft, Service Health Status, status.cloud.microsoft — service degradation notice for Microsoft 365 (Business or Enterprise) and consumer products, page updated 13 Nov 2026 03:07:45 UTC; opened 13 Nov 2026.
- Microsoft 365 Status (@MSFT365Status), post on X, 13 Nov 2026 1:54am MYT, reference EX1550543 — opened 13 Nov 2026.
- ABC News, Microsoft platforms including Outlook suffer widespread outages, 13 Nov 2026 — Microsoft’s statement on elevated load, the affected service list, and the recovery language; opened 13 Nov 2026.
- Central News Agency (CNA), citing Reuters, 微軟365、Outlook等服務當機, 13 Nov 2026 — Downdetector figures at 2:07pm US eastern, and the OpenAI and Anthropic connector statements; opened 13 Nov 2026.
Keep ahead of the next disruption. BD Media publishes Malaysian business-technology reporting every weekday — subscribe to the BD Media briefing and get the analysis in your inbox.
