Technology 149 sources · over 7 days Latest coverage 19 Nov 2026, 5:45 pm UTC

Microsoft Blames Months-Old Bug for Global 365 Outage That Knocked Out Outlook, Teams and Copilot

Microsoft's own post-incident report says a flaw dormant since July, not a cyberattack, took down core workplace tools for almost a day, raising fresh questions about how much the world relies on a few providers.

By Hushread Stories, from 149 outlets · First published 19 Nov 2026 · Chapter 4 of 4

Earlier: Microsoft 365 outage traced to a bug dormant since July

In brief

  1. A software bug introduced in a July update to Microsoft's gateway servers lay dormant for months before triggering the outage on 12 November.
  2. At the peak, more than 80% of the networking gateway servers in front of many Microsoft 365 services were unresponsive.
  3. Outlook, Teams and Copilot were disrupted for close to 24 hours, with problem reports coming from 72 countries.
  4. Microsoft is offering affected business customers a credit of 50% of their subscription price, which must be claimed by 31 December 2026.
  5. The outage came amid a string of US tech failures, but Microsoft has said it was not caused by a cyberattack.
Microsoft Blames Months-Old Bug for Global 365 Outage That Knocked Out Outlook, Teams and Copilot
Source: The Verge - All Posts Tech

Timeline · 10 moments

10 moments Open the full timeline →

Nadella calls outage a perfect storm as shares slide

CNBC Technology ↗

Executives tell Congress outages stem from rushed security patching

Lifehacker Tech ↗

Outage seen alongside FAA and bank failures in US

CNN.com - RSS Channel - App International Edition ↗

Hackers exploit outage for credential phishing

Computerworld ↗

Over 80% of gateway servers down at peak, Microsoft says

Business Line - Home ↗

Microsoft offers 50% subscription credit to affected firms

tech ↗

Bug lay dormant for months before load surge triggered it

The Next Web ↗

Microsoft releases preliminary report blaming a July bug

Computerworld ↗

Problem reports pass 43,000 across 72 countries

Fortune ↗

Microsoft 365 outage begins, hitting Outlook, Teams and Copilot worldwide

The Verge - All Posts Tech ↗

How it started

On Thursday 12 November, Microsoft 365 began failing for users around the world. Outlook, Teams and Copilot were among the services hit, and by the next morning Fortune counted more than 43,000 problem reports from 72 countries.

The Verge described the start of it from inside the company. At 9AM on the Thursday, it reported, everything seemed normal as Microsoft employees arrived for a regular day of emails and meetings. Barely an hour later, one of the company's worst outages in history was under way.

Microsoft's first public explanation pointed to a technical issue linked to a network capacity upgrade. It said the problem was not a cyberattack and not related to AI agents.

How it unfolded

Repairs did not go smoothly. Microsoft tried a series of fixes over many hours, but planned repairs did not fully work, and services recovered only slowly through the following day. Mashable noted that the disruption stretched across two days starting on 12 November.

The fuller picture arrived on Monday 16 November, when Microsoft released a preliminary post-incident report. Computerworld said the report identified multiple causes, including an admission that a bug had gone undetected. BleepingComputer reported that a software bug introduced in July triggered a cascading failure that left more than 80% of the networking gateway servers unresponsive.

The Next Web said the bug lay dormant for more than three months before a surge in load and a drop in capacity set it off. GeekWire added a twist: Microsoft's own safety systems helped turn the dormant flaw into a global outage, and the company blamed what it called "thundering herds" of retrying traffic. Ars Technica summarised it as three small failures combining. Business Line reported that automated traffic rerouting, server capacity constraints and testing gaps all played a part.

On Tuesday, Business Line reported that the explanation came from the company's Executive VP of Copilot, who said three factors together caused the gateway outage. By Wednesday, CNBC reported that CEO Satya Nadella had called the incident a "perfect storm", and noted that Microsoft shares had dropped nearly 9% since closing at a record high on 4 November.

The outage also gave attackers an opening. Computerworld reported that hackers exploited it to run a credential phishing campaign, the day after the disruption.

Where it stands

Services are back, and Microsoft's account of the cause is now fairly detailed, though it is still labelled preliminary. French outlet tech reported that affected companies can claim a credit worth 50% of their subscription price, and have until 31 December 2026 to do so.

The outage landed in the middle of a wider run of failures in the US. CNN listed email going down, then flights being grounded by the Federal Aviation Administration, and Mashable added a Chase bank outage caused by a software problem. The Independent reported that President Donald Trump brushed off the series of outages in remarks on Tuesday.

Whether these events are connected is disputed. Lifehacker reported that Wall Street executives and the NSA told Congress the outages came from companies rushing to patch security flaws found by AI models, not from hackers. New Scientist cautioned that speculation about an AI hacking spree or a coordinated foreign attack was wide of the mark. For Microsoft's own outage, the company's explanation points to an internal software defect.

What to watch

The post-incident report is preliminary, so a final version with more detail on the testing gaps and on what Microsoft will change is the obvious next step. Customers eligible for the 50% credit also have until 31 December to make their claim.

The larger argument is about dependence. The Hindu and CNA both ran commentary on the dangers of centralised digital systems, and CIO asked why customers do not simply leave Microsoft. Whether the outage prompts any real change in how companies and governments spread their risk is still an open question.

Spotted an error? Email [email protected] and we'll fix it.

More in Technology

All →