The key facts

  • The outage began on Thursday with Exchange Online and affected Outlook, Teams, Copilot and SharePoint in more than 70 countries
  • Microsoft attributed the cause to a technical problem during a network capacity update, ruling out a cyberattack and AI
  • The company apologized and promised to publish a full report once the investigation is complete

What happened

A global outage of Microsoft services began on Thursday with Exchange Online — the platform that provides corporate email through Outlook — and lasted nearly 24 hours. The company first confirmed problems with mailbox access for users in North and South America at 17:54 GMT on Thursday, and soon warned that the outage might be affecting Microsoft 365 worldwide. In addition to email, functions of services dependent on Exchange Online were disrupted: Teams calendars, Copilot and SharePoint.

Cause and recovery

Microsoft said the outage was caused by a technical problem unrelated to either a cyberattack or artificial intelligence or AI agents. According to the company, a latent defect emerged during an earlier network capacity update and manifested after a recent change to the network. Recovery proceeded gradually due to the scale of the load: engineers checked stability at each stage, which, the company acknowledged, took longer but ensured a reliable restoration. By Friday evening, Microsoft reported 'encouraging signs of recovery' in the Asia-Pacific region and improvements in North America and Europe, but did not give a timeline for full restoration. The company apologized to customers and partners and promised to publish a full report once the investigation is complete.

Scale of the consequences

The outage affected more than 70 countries, and Microsoft products are used by more than 430 million people worldwide. Institutions were hit, from the government of Japan and the Supreme Court of the Philippines to Australian intelligence; in North America, problems arose at railways, hospitals and courts. In the United Kingdom, part of the NHS, schools and police were affected: cybersecurity minister Liz Lloyd said the impact on critical services, including 999 calls, was limited, and the transport network was not affected. In Canada, the national railway network warned that passengers might not receive email notifications about their trips. In the United States, according to House of Representatives staff, the outage derailed an entire day of legislative work. Problems also arose at the government of Singapore, among delegates at the COP31 climate summit in Antalya, and at small businesses: South African winemaker Peter Pentz said his production had effectively come to a halt.

Dependence on cloud services

Experts called the incident a reminder of how dependent organizations are on a small number of large technology providers. Rob Demaine, head of the company e2e-assure, called the outage a 'sharp reminder' of dependence on a few major suppliers. Matt Saunders of the consulting firm Adaptavist noted that the classic redundancy strategy of using a second service is in practice almost unfeasible, since it is difficult for businesses to operate without Microsoft tools, and urged organizations to prepare contingency plans for such outages in advance.