Skip to content
10 ways the 2026 Microsoft outage exposed organizations with no fallback

10 ways the 2026 Microsoft outage exposed organizations with no fallback

Qz •Colleen Cabili • November 13, 2026

From sovereignty to schools to a Kent removals firm, 10 lessons show what organizations lost by keeping nothing outside Microsoft's platform.

Most offices treat email and calendars as background equipment that staff notice only when it stops working. When Microsoft $MSFT 's email and scheduling service failed, ordinary work stopped with it in governments and businesses on several continents. Reuters reports that the failure reached public agencies in the U.S., Japan and Australia. Private companies carried the same exposure on the same day, with 3.9 million businesses worldwide running on that Microsoft suite, according to ABC News . Staff could not send messages, check meetings or get answers. The fault did not clear quickly, so organizations had to improvise around it for hours instead of waiting it out. Calendars, messages and decisions had all depended on a single service.

What should worry organizations is that Microsoft keeps failing. Microsoft 365 had already gone down in October 2025 and again in January 2026, according to CBC News . Reuters reports that a separate failure of Microsoft 's Azure cloud service in October 2026 also hit the company's productivity software. Organizations should plan for breakdowns as a recurring cost of relying on these tools. Investors barely reacted to the latest one, marking Microsoft shares down 1.7% before regular trading began, a modest drop for a stock worth roughly 8% more than at the start of the year. They treated the failure as a small matter for Microsoft , even though its customers lost hours of work.

The Guardian set out the lessons organizations and governments can draw from the long Microsoft service failure. Around 18 hours in, the outage was still continuing. The lessons cover the push for digital sovereignty, emergency security workarounds that can stay in place after the crisis, schools cut off from email, and a small removals firm whose loaded vans could not leave. Each one asks what an organization could still do once the platform beneath it stopped. Each also asks what the organization paid for keeping nothing outside that platform. These are questions preparation, and they apply to every organization that uses the platform. Microsoft 's repair does not settle them, and they remain open after the email returns.

Single-supplier dependence

A continuity plan fails its first test if staff cannot reach it when the main system goes down. Organizations whose backup also ran on Microsoft $MSFT failed that test. "A continuity plan that relies on the same platform that has just failed is not a plan," Steven Watt, a professor of practice in cybersecurity at the University of St Andrews, told The Guardian. In Australia, the Australian Federal Police and the Australian Security Intelligence Organisation both confirmed they were affected, according to ABC News . Even agencies charged with national security depended on the platform that failed. A plan is only useful if some part of it runs outside the supplier that has just failed.

Australia's government chose this dependence with the risk already named. The Critical Infrastructure Security Centre released a risk review on Nov. 2, 2026, ABC News reports , less than two weeks before the outage. Months earlier, a Digital Transformation Agency review made public in February 2026 had judged the overall risk of depending on one vendor to be "medium." The government kept buying anyway, giving Microsoft federal contracts worth more than $1 billion this year, including a $720 million deal with Defence. Nobody in government could call the danger a surprise.

When Microsoft failed, Australia responded by drawing closer to Microsoft rather than turning to a second supplier. The government activated an agreement with Microsoft , signed in June 2026, on the day of the outage, according to ABC News , and Affairs Minister Tony Burke says the government used it to make sure Microsoft gave Australian interests suitable weight in its response. That arrangement may win Australia more attention during Microsoft 's recovery work, but Microsoft 's platform remains the only one the government can run on. Even with that attention, the government still has to wait for Microsoft to repair its own service, with no other system in place during the wait.

For any organization, a closer relationship with its main supplier is no substitute for a second one.

Digital sovereignty and vendor lock-in

Digital sovereignty often sounds like a debate national pride, but during the outage it became a practical question of whether a government can keep working when a foreign supplier's service stops. Dame Chi Onwurah, chair of the Science and Technology Committee in Parliament, argues that relying on a few U.S. providers blocks homegrown alternatives, according to The Guardian . On that view, the dependence feeds itself: the more the state relies on a few suppliers, the less room any rival has to grow. Onwurah treats sovereignty as a matter of keeping services running, and she said the task is working out "how can we protect against vendor lock-in or achieve the resilience and digital sovereignty the UK needs." Before a government can reduce its exposure, though, it has to be able to measure it.

The British government could not give Onwurah's committee a basic account of that exposure. In its response, it could not say how much it spends on IT in total, according to The Guardian . Without that figure, it has nothing to set against the cost of an outage or of an alternative. It also could not say who its suppliers are. That gap matters because public bodies such as the NHS and the Metropolitan Police were among the organizations the outage hit, according to Reuters . Nobody can plan a fallback for a supplier they have not listed, so sovereignty begins with an inventory, however unglamorous that work is.

Keeping control over where the software runs made a difference during the outage, even without a homegrown replacement. Organizations that hosted Microsoft $MSFT 's software on machines they controlled themselves got off somewhat lighter, according to Reuters . A government can take that step now, without waiting for a new supplier to appear. Josh Gilbert, a market analyst at eToro, says repeated failures give IT departments cause to hesitate before moving still more of their work to a single company. The realistic goal is to stop adding to the dependence rather than to tear it out at once. For a government, the test is whether it can answer the two questions it could not answer for the committee: what it spends on IT and with whom. Only then can it decide what to keep in-house.

Europe cannot safely rehearse a deliberate, politically motivated cutoff of American office software, because the only real test would be the cutoff itself. The accidental failure gave organizations a version of that test without anyone having to choose it. Dario Maisto, a principal analyst at Forrester, calls the outage a natural experiment in how Europe would cope with a politically motivated cutoff. Mail and Copilot AI tools are the services at the center of that scenario, and the real failure struck that same mail layer. Microsoft $MSFT tied the damage to Exchange Online and to the applications that connect to it, according to CBC News . For one day, each organization could see where its dependence on that single mail layer actually ran.

The rehearsal was gentler than the real thing would be. Maisto himself treats the failure as an accident rather than a government decision, so it can only approximate what a politically ordered loss of mail and AI tools would do. That difference matters most for how a crisis ends. In an accident, the vendor works to bring service back. In a deliberate cutoff, nobody would be racing to restore anything, and the disruption would last as long as the political dispute behind it. An organization that got through this outage by waiting has learned the least, since waiting would not work in a deliberate cutoff.

Whatever stopped during this accident marks a place where an organization's work could not go on without the platform, and finding those places is the purpose of a rehearsal. Organizations should treat those breaking points as the minimum damage. Anything that broke under an accident would break sooner, and stay broken longer, under a deliberate cutoff.

The test has been run once, for free, and the open question is whether organizations fix what it found before anyone runs it on purpose.

Security controls and emergency changes

During the outage, many organizations lost the software that guards their data along with their email, and the shortcuts staff took to keep working may stay in place long after service returns. CoreView, a security firm, advises organizations to confirm that data protection works again and to review the fixes they rushed into place, according to The Guardian . Simon Azzopardi, CEO of security firm CoreView, says that in many cases IT teams were without their security and compliance software as well as their inboxes. A company could stay open for business and go unguarded at the same moment, and a restored inbox proves nothing restored defenses. The larger hazard is whatever staff improvised to get through the day. "Review every emergency change made during the outage, because workarounds made under pressure tend to become permanent gaps," Azzopardi said. Organizations do not settle the security bill for the outage on the day email comes back. They keep paying it until someone goes back to check.

An organization can pass a security audit and still have work to do on planning for a stoppage. In 2024, the Office of Inspector General for the Federal Reserve Board and CFPB found that the selected Microsoft $MSFT 365 security controls it tested at the Consumer Financial Protection Bureau were operating effectively, according to Federal Reserve OIG . The same January 2024 audit found that the agency could strengthen certain contingency planning controls, though the details of that weakness are not public. Contingency planning covers the moment the normal system fails, which is when staff make the kind of workarounds Azzopardi warns . That audit looked only at one agency's own system in 2024, not at this outage, but it shows that strong security controls and a sound plan for a breakdown are separate things to check.

Every organization should plan for losing its protections as well as its inbox and decide in advance which shortcuts it will allow. The cleanup afterward should follow a checklist of specific controls rather than a general sense that things are back to normal. Sensitivity labeling, which marks data as sensitive so it can be protected, is one control CoreView says to confirm. An organization that skips that review may carry an emergency setting into an ordinary week without anyone noticing.

Delayed email and attackers

Delayed email deserves extra suspicion after an outage, even one that no attacker caused. Microsoft $MSFT found no sign of malicious activity behind the failure, according to CNBC . Security specialists still treat the confusion that follows a disruption as an opening for criminals. Simon Azzopardi, CEO of security firm CoreView, told The Guardian that "attackers routinely use outages as cover." On a day when staff receive almost everything late, they have little reason to question one more late message, and a fraudulent request becomes harder to spot. Until the backlog clears, people should read delayed mail with more doubt than usual.

Microsoft 's effort to save delayed mail also lengthened the stretch in which late messages keep arriving. The company extended its retry window for delayed email to 48 hours, CNBC reports , so it could keep attempting delivery during the recovery instead of giving up. Azzopardi says Microsoft cautioned that Safe Links, its tool for screening web addresses in mail, may not have worked while the disruption lasted. While the outage went on, late messages could therefore reach staff with links that one of Microsoft 's own filters had not screened. Keeping mail moving was the right decision for business continuity, but it means delayed messages, and the doubt they call for, can continue into the recovery.

The practical lesson is to keep a way to check messages that sits outside the hub. Staff who get an unexpected request or link in delayed mail should confirm it through a channel that does not depend on the same inbox, such as a phone call to the sender. Organizations that have no such channel discover the gap in the middle of an outage.

Microsoft decides when service is back, but the exposure lasts until staff have opened the last delayed message with care.

Fallback channels in public services

A public service could keep working through the outage only if its backup needed nothing from the platform that failed. On the same day, NHS staff went back to paper and the city government in Tampa, Fla., kept its phone lines and 911 running, while police in the U.K. could not take reports online, according to The Guardian . A pen and a phone line work without an email-and-scheduling hub, but an online reporting form does not. Tampa officials kept city offices open on normal schedules, so residents who came in that day got ordinary service instead of a closure notice. The services that kept going were not the most technically advanced. They were the ones where someone had kept a working channel off the platform.

A service that runs only online fails hardest when members of the public are trying to report a problem. "We are currently unable to receive complaints, whistleblowing and malpractice submissions," said Ofqual, England's exam regulator. People with evidence of cheating or misconduct in exams had nowhere to send it, and those reports carry the most risk when they go unheard. When a regulator loses its only intake channel, it is left with a gap in oversight for as long as the platform stays down.

Keeping a service running by hand did not let it get through the outage free. ABC News reports that Ashwin Pal, a Risk Advisory partner at RSM Australia, sees the catch-up effort after an outage, more than the initial downtime, as what costs money. A working fallback keeps people served during the outage, but it postpones part of the cost rather than avoiding it. Public services need a channel off the platform to keep serving people during an outage, and they need a budget for the cleanup that follows.

The outage reached every level of education at once, from the classroom up to the bodies that oversee it. Schools and universities lost email services, according to The Guardian , and the disruption also hit Ofqual, England's exam regulator, and Ofsted, the schools inspector. The inspector and the schools it judges were caught in the same outage, so no part of the system could cover for another. ABC News reports that long-distance schools shut for the day. For pupils taught at a distance, the outage cost a full school day rather than slowing one down.

For schools, Microsoft $MSFT 's platform holds much of the coursework as well as the mail. The education features that may have been affected included assignments, which teachers set and pupils hand in, along with Reading Progress and Speaker Progress, tools that pupils use to practise reading and speaking. That puts pupils' work at risk, not only staff messages. A lost inbox slows the paperwork around a lesson, but when the coursework sits on the same service, teaching shares its single point of failure.

The loss that matters most for trust in the system was the set of routes for reporting wrongdoing. "We are currently unable to receive complaints, whistleblowing and malpractice submissions," Ofqual said. Those three routes are how pupils, parents and staff raise concerns unfair treatment or cheating in exams, and all three closed together.

Ofqual told people with urgent enquiries to use the telephone, the one route that did not depend on the platform that had failed.

Small-business knock-on effects

A small business can lose a working day to an email outage even when its own work has nothing to do with software. In Kent, England, crews from Stranks Removals & Storage sat idle and a house exchange stalled when emails couldn't be sent, according to The Guardian . The firm had two crews out waiting for keys. Neither could do any paid work until those keys arrived, and the firm had no way to speed that up. Its people were in place and ready to move furniture. What they lacked was a confirmation that depended on someone else's inbox.

The hold-up began on the client's side of the deal. A client could not exchange on a property because they could not send emails. That left the firm's day dependent on a step in a property sale it had no control over. A small firm's exposure to an outage therefore extends to everyone it waits on, including buyers, sellers and whoever has to confirm a deal before keys change hands. Switching its own email provider would not have put those crews back to work. The dependence sat in tools the firm never chose.

For a business that works to fixed dates, a lost day can also put later bookings at risk. One delayed exchange involves a move planned to start Nov. 23, and the hold-up could push that date back. Businesses at the end of the chain pay for an outage like this, even though they have the least power to prevent it and the least room to absorb it. The practical step for a small firm is to agree in advance on a route outside email with the people a deal depends on, such as a phone confirmation or a paper backup. Then the clients and the firm can still complete a deal when one platform fails.

Dependency chains inside Microsoft's products

Microsoft $MSFT 's assurance that Teams stayed up was accurate in a narrow sense and misleading in a practical one. During the outage, the company said the core functions of Teams were working, The Guardian reports , while the Teams calendar and Copilot features that rely on Exchange Online had failed. Teams was still among the products affected, along with Outlook and Copilot, according to CBC . For the people using it, the line between working and broken ran through the inside of Teams, along whatever each feature needed from Exchange Online. The name on a product tells a customer little how it will behave when the email-and-scheduling hub beneath it goes down.

Staff could still talk to one another, but only in the parts of Teams that needed nothing from that hub. Standard Teams channels kept working, so colleagues could still message one another there. The newest feature in the suite is only as available as the mail, calendars and files it reads. Inside Teams, Copilot could not answer anything that touched Outlook, calendars or Microsoft 365 files, and CBC lists Copilot as an affected product in its own right. Organizations also could not create a Teams live or virtual event, which left them without the tool for briefing a whole workforce at once, just when a crisis called for one. Microsoft 365 failed as one system built around a shared hub, and spreading work across its apps adds features, not redundancy.

Switching from Outlook to Teams during an outage moves work onto another part of the same system whose hub has just failed. "A continuity plan that relies on the same platform that has just failed is not a plan," said Steven Watt, a professor of practice in cybersecurity at the University of St Andrews. Teams lost every feature that leaned on Exchange Online, so a plan that amounts to switching to Teams fails his test.

Real redundancy needs something that draws nothing from Microsoft 's hub, and no organization can get it by buying more Microsoft products.

Slow recovery and an unknown cause

Microsoft $MSFT restored its service in stages, and some users were still stuck after most were working again. In a status update, Microsoft said the pile of waiting mail had shrunk to near its usual size, but it cautioned that some people could still drop in and out of their accounts or get messages late until the recovery was finished, according to a Microsoft status update reported by The Guardian . An organization could not treat that all-clear as the end of the disruption. For a business, the end of an outage is not a single moment it can plan around. It is a stretch of uncertain days that lasts until the supplier's recovery reaches its own staff, and those days carry a cost on top of the hours of outright failure.

The unexplained part of the cause is narrower than it sounds. ABC News reports that the company traced the stoppage to a process on its servers that sets up secure connections and became overwhelmed, which left the servers unable to handle requests. The breakdown happened inside Microsoft 's own systems under heavy demand, and the company said it saw nothing to suggest an attack. Knowing that the process was overwhelmed is still not the same as knowing why, and the reason is what bears on whether it happens again. Without that reason, customers cannot judge the odds of another failure. They can only decide how much of their own work will be waiting when one comes.

Some users lost access to far more than their inbox. ABC News reports that with Outlook down, some users could not receive two-factor login codes and were locked out of other services. Recovery meant undoing a chain of lockouts, not repairing one service. Customers got no protection from the size of Microsoft 's operation either, since the fault sat inside the company's own servers, and Microsoft had to bring service back in Europe, Japan, North America and Asia Pacific, according to The Guardian . "If all of them are failing, nothing can be done," said Adel Toosi, an associate professor at the University of Melbourne.

A customer cannot speed up a supplier's recovery. The only part an organization controls is how much of its work has to wait for it.