/
Microsoft says Outlook, Teams, Copilot and other Microsoft 365 services are fully restored after an outage across at least 74 countries that lasted over 24 hours. It blames a network capacity upgrade, not a cyberattack.

Microsoft has said that Outlook, Teams, Copilot and the other Microsoft 365 services knocked out in a global outage are now fully restored. The disruption reached at least 74 countries and, by the time the company declared recovery, had gone on for more than 24 hours.
The company said the remaining regional clusters have been brought back and the queue of stuck messages has been cleared. Some mail sent during the outage may still arrive late, as the backlog works through the system.
The Microsoft 365 suite is the backbone of daily work for businesses and public bodies, and the failure landed on its most widely used parts. Outlook users had trouble with email delivery and with reaching their inboxes. Teams, the company's collaboration platform, was disrupted, and so was Copilot, its AI assistant. SharePoint was also affected, according to the service problems described as the outage unfolded.
The geographic spread was wide. Problems were logged in at least 74 countries, and the effects were not limited to office workers. Governments, courts, healthcare services, businesses, airlines and banks in many places all had services disrupted. Several national governments, Japan's among them, were caught up in the outage.
For organisations that run email, meetings, document storage and AI tools through one provider, a fault at that provider can stop several functions at once. That is what happened here: messaging, video calls and the assistant tools that sit on top of them were all unreliable together.
Microsoft's own account of the incident shows a long recovery rather than a quick fix. While the fault was still unresolved, the outage had already run for at least 18 hours, and services such as Teams, Outlook and Copilot were all still affected. Later, once the company announced that service had returned, the total disruption was described as having lasted more than a day.
The recovery was staged. Microsoft worked through the affected regional clusters one by one, and its final confirmation came only after the last of them was restored and the backlog of queued messages had been cleared. Only then did the company state that all affected Microsoft 365 services were back to normal.
Microsoft also apologised. The company said it was deeply sorry for the outage and the disruption it caused to customers, among them the healthcare services and public institutions that depend on its tools.
Users have been warned that delayed messages should still be expected for a time. Mail that was queued while the fault was active is being delivered as the system clears it, so an email that appears hours or even a day late does not by itself point to a new fault.
The cause matters because outages of this size often prompt a first suspicion of hostile action. Microsoft ruled that out. The company said the incident was not a cyberattack and was not connected to AI agents.
Its explanation is technical and fairly narrow. Microsoft said the problem was triggered by an issue during an earlier network capacity upgrade. After a recent network change, servers became unexpectedly unavailable, and that unavailability spread into the services people use every day. In plain terms, work meant to add capacity to the network ended up taking part of it out of service.
Microsoft has not, in the statements available, said who within the company was responsible for the change, and it has not described what went wrong in the upgrade beyond the outline above. No regulator has been cited as commenting on the incident in the material available.
The scale of the outage helps explain why the cause is being watched closely. Courts, hospitals, airlines and banks are among the organisations that rely on cloud-hosted email and messaging, and when those tools fail they have to fall back on other methods or wait. Microsoft's confirmation that the cause was internal and not malicious means that, on the company's account, no customer data breach is part of the incident. Microsoft has not offered an assessment of that kind in the statements available, so customers wanting certainty on that point will be looking for more detail from the company.
For individual users and IT administrators, the practical position is that service has returned and the remaining task is the backlog. Outlook mailboxes should be receiving mail normally, with the possibility of late arrivals from the queue. Teams and Copilot are listed by Microsoft as restored.
Anyone still seeing sync or sign-in problems after the company's all-clear is in a different position from those caught in the main outage, since Microsoft says the affected regional clusters have all been recovered. Such cases would be worth reporting through the company's usual service channels, because they would not be explained by the incident Microsoft has now closed.
Organisations that had to improvise during the outage, by moving meetings to other tools or relaying messages by phone, can now return to their normal systems. Microsoft has not published a separate timetable for compensation or service credits, and none is mentioned in the material available.
Microsoft has not yet set out a date for a fuller account of what went wrong in the network upgrade. No time has been given for a final incident report, and the company has said nothing further on how it will prevent a repeat. With the services restored and the queue of delayed messages still draining, the next word from Microsoft is expected to be a closing update on the backlog, though no time for it has been announced.
About the Author
Staff Writer
Amita Sahani writes for Wirony. Every story under this byline is reviewed and approved by an editor before it is published, and corrections are made openly.
What did you think?
Italy's data protection authority, the Garante, has fined Meta Platforms Ireland €805 million over a Facebook Android feature that accessed photos and videos without specific consent. Meta has 60 days to change the workflow and may appeal.
Subhash Gupta5 min
Grand Theft Auto VI launches on Thursday, 19 November, on PlayStation 5 and Xbox Series X|S, at $80 standard and $100 for the Ultimate Edition. Preloading is under way, the game needs at least 210 GB, and no PC version is announced. Reports differ on preload dates and some figures.
Subhash Gupta4 min
Outlook, Teams and Copilot stayed disrupted into 13 November after Exchange Online servers failed health checks. Microsoft extended its email timeout to 48 hours so delayed mail is not bounced and reports phased recovery. Hospitals, airlines and governments were hit. It found no sign of an attack.
Amita Sahani4 min
President Volodymyr Zelensky said on 12 November that a speed cap of about 161 km/h on Starshield, the military version of Starlink, has returned and is hampering interception of Russian jet-powered drones. Musk says the US military decides; Washington has not accepted responsibility.
Subhash Gupta4 min
Google announced two new eighth-generation TPU systems on 9 November. TPU 8tx is for confidential inference, with up to 2x lower latency than TPU 8t, and TPU 8i Max doubles compute, SRAM and HBM bandwidth over TPU 8i. The post gives no prices, memory sizes or availability dates.
Prateek Rawat4 min
China's Foreign Ministry said on 9 November that the Chinese spy allegation linked to Nvidia's chief executive is unsubstantiated and a smear. Spokesperson Guo Jiakun gave one-line answers on a reported 600,000 Nvidia chips in Hong Kong and on US curbs. Nvidia's board has opened an inquiry.
Amita Sahani4 min
Nvidia's board announced on 8 November an independent investigation, overseen by independent directors and outside counsel, after reports that CEO Jensen Huang was the target of a foreign intelligence effort. It has no reason to believe security or secrets were compromised; Huang will cooperate.
Subhash Gupta4 min
Rockstar says GTA VI pre-loading starts at midnight on 10 November, two days before the 12 November date given at pre-order. The download is up to 212 GB, boxed copies reach shops on 12 November and the game launches on 19 November.
Prateek Rawat4 min
Canva said on 4 November that 1,271 people, 22% of its staff, will lose their jobs or change roles or locations. Leavers get at least six months' pay, a year of healthcare and US$10,000 start-up grants. Product, Growth and Platform merge.
Amita Sahani4 min
On the night of 1 November, Ukraine's Air Alert app sent false warnings of radiation and a nuclear attack; about three hours later Russia's Emergencies Ministry app sent a false nuclear alert to 33 regions. Zelensky told a US broadcaster Ukraine hacked the Russian app in reply. Moscow names none.
Prateek Rawat6 min
Binance says hackers the FBI links to North Korea took about $6.3 billion on 29 October after months of work: a tampered installer in June, a fake recruiter and coding test in September, and an internal AI tool on 20 October. About 98% of losses are replaced.
Prateek Rawat5 min
DeepSeek plans at least 160,000 Huawei Ascend 950DT chips at a roughly 1 GW Inner Mongolia site and reportedly aims to start pre-training V5 on them from late 2026, say unconfirmed reports. Arm shares fell about 14% in two sessions to 2 November after Intel's weak outlook.
Subhash Gupta5 min