As of about 5:00 PM IST, Saturday 14 November 2026. Based on Microsoft's Saturday restoration notice, the company's Friday explanation, and reporting by CNBC, The Independent and Deutsche Welle.
Microsoft said on Saturday that it had finished fixing the underlying problem behind a global outage of Microsoft 365 and had restored service to the remaining affected customers and regions. Outlook email, Teams and the Copilot assistant were among the products that failed, in a disruption that lasted more than 24 hours and reached governments, hospitals, courts and businesses on several continents.
The company has said the incident was a technical failure, not a cyberattack and not the work of artificial-intelligence agents. That assurance matters, but it does not shrink the practical lesson. For more than a day, a large share of the world's office work depended on a single vendor's network coming back.
A disruption that outlasted a working day
Problems began on Thursday afternoon, 12 November. CNBC reported that connectivity issues started shortly before 1 p.m. Eastern Time. The network diagnostics firm Ookla, cited by Deutsche Welle, put the start around 17:25 UTC. From there the outage followed the clock: it hit Thursday evening in Europe, ran through Friday's business day in Asia and Europe, and was still being unwound on Friday morning in the United States.
Users reported that they could not send or receive Outlook mail, and that calendar, SharePoint and Copilot features failed. The Independent said Downdetector recorded problems from more than 70 countries and territories, with more than 130,000 Outlook reports in the United States at the peak on Thursday night. Deutsche Welle, also citing Downdetector, put the worldwide total of problem reports above 7.4 million. Those figures are not a census of unique users, but they describe a failure that was both wide and deep.
On Saturday, Microsoft said it had "completed our efforts to address the underlying issue of the incident and restored service to the remaining impacted customers and regions," pointing administrators to incident MO1550821. The Independent published that confirmation on Saturday morning in the United States. Deutsche Welle described the episode as one of the longest in Microsoft's history, and the second such event this year.
What Microsoft says went wrong
Microsoft's account sharpened over two days. Early updates on Thursday said a number of servers had become "unexpectedly unavailable" after failing automated health checks, which raised the load on the remaining infrastructure and produced connectivity errors. Engineers tried restarting servers, shifting traffic and reversing recent configuration changes. Some of those steps did not deliver a lasting recovery.
On Friday the company apologised and offered a clearer cause. "We are deeply sorry for the disruption this incident has caused our customers and partners. We know the trust they place in our services and take that responsibility extremely seriously," a spokesperson told CNBC. Microsoft said the incident "was introduced by an issue during an earlier network capacity upgrade that surfaced following a recent network change," and that it was "a technical issue not related to cyber or AI / agents."
Officials and commentators had already asked whether the outage was malicious, or somehow tied to autonomous software. Microsoft has said it detected no malicious activity. A detailed post-incident report has not yet been published, so the upgrade narrative should be treated as the company's current explanation, not a finished investigation.
A slow, staged recovery
Microsoft did not bring the world back online with a single switch. CNBC reported that engineering teams took affected systems offline for fixes and returned them in stages, and that Asia-Pacific services were among the first to start returning to normal. By Friday morning, the company said email queues were "approaching normal levels," then later described improving delivery performance. It also extended the timeout window for delayed messages to 48 hours, so mail that had piled up would keep trying to send rather than being dropped.
That choice underlines what an email outage actually costs. The risk is not only that a message fails in the moment, but that a day's invoices, prescriptions, court notices and customer queries sit in a queue. Gradual restoration is slower to announce than a clean "all clear," but it is how operators try to avoid a second collapse when deferred traffic returns at once.
Who was forced onto paper and workarounds
The Independent reported that the outage reached governments, courts, healthcare services, businesses, airlines and banks. Officials in Japan, Singapore and Australia said some of their operations were affected. In Britain, NHS England said staff were using long-standing fallbacks, including paper records. The National Pharmacy Association warned of longer waits and delays to some prescriptions, because doctors could not always email them through. Police forces in England and Wales said they could not receive crime reports online. Government cyber minister Baroness Lloyd of Effra said the failure mostly hit back-office functions, while acknowledging "limited reports of impact on frontline service delivery."
Companies felt it too. CNBC reported that Swiss luxury group Richemont slightly delayed its earnings, and that electric-truck maker Workhorse Group pushed a planned results release from 12 November to 16 November. Air India said it had temporarily rerouted customer email and that flight operations were not affected. Some users also reported problems with other internet services on Thursday; Amazon Web Services said its own systems were operating normally and that an unrelated disruption was producing inaccurate reports. The Independent added that other AI chat services, including OpenAI's ChatGPT and Anthropic's Claude, were disrupted as well. It did not establish that those failures shared Microsoft's cause.
The real issue is concentration, not mystery
Once a cyberattack is ruled out, the remaining problem is architectural. Email, video meetings and workplace AI have been bundled into a handful of cloud platforms. When one of those platforms stumbles, remote teams that organise their day around those tools discover how little of their work can move without them. James Owen of the consultancy Control Risks told CNBC that organisations had become "increasingly dependent on a smaller segment of technology and infrastructure," and that the incident underlined the need for recovery plans that do not assume the internet, or a single supplier, will be available.
That is a more useful conclusion than a search for villains. It also sits alongside a broader debate about whether the industry's infrastructure build-out is running ahead of resilience, a theme we have examined in the context of debt-funded data centres and the AI boom. Copilot was one of the products that went dark, but the more basic failure was email. The organisations that coped best were those that still had a way to work when the primary system did not, including manual procedures and backups they could actually restore.
What to watch
Microsoft still owes customers a full incident report: what, exactly, went wrong in the capacity upgrade, why the first fixes failed, and what has been changed so that a health-check failure cannot again take down mail for a day. Until that report appears, the Saturday restoration is the end of the outage, not the end of the story. For any organisation that ran on Outlook or Teams this week, the practical test is simpler. If the same services vanished on Monday morning, how long could the work continue?
Sources
- CNBC: Microsoft apologizes for massive outage that took down Outlook and other 365 services
- The Independent: Microsoft confirms services fully restored after major global outage
- Deutsche Welle: Microsoft says all services working again
Featured image: Building 92 of the Microsoft Redmond campus, 18 June 2015, Jiaqian AirplaneFan, CC BY 3.0, via Wikimedia Commons (cropped).