Microsoft announced on Saturday that services were fully restored following a massive, hourslong global outage that impacted users across more than 70 countries and territories, according to The Associated Press.
The disruption affected major Microsoft 365 services including Outlook, Teams, and the AI assistant Copilot, causing widespread difficulties for governments, courts, healthcare services, businesses, airlines, and banks. The company explicitly denied that the incident was related to artificial intelligence or cyberattacks, attributing the disruption instead to a technical issue stemming from a recent network capacity upgrade.
Problems initially emerged midday on Thursday in the United States, with The Associated Press reporting that online tracker Downdetector received outage reports from 74 countries and territories within an hour. According to The New York Times, hundreds of Microsoft engineers and top executives spent Thursday night logged into a video conference triaging the crisis while dealing with a server failure that left even internal staff unable to use email.
Scott Guthrie, the senior-most executive overseeing the company’s computing infrastructure, proposed fixes during the call as participants debated how to prevent a stampeding herd of new demands from overwhelming systems as they came back online.
The technical failure caused cascading effects worldwide as servers unexpectedly became unavailable during automated checks. For Outlook users, many were unable to send or receive emails or load their inboxes altogether. In Britain, the outage halted a trial in Manchester and prompted the National Health Service and the National Pharmacy Association to warn of potential delays in getting medication and emailing prescriptions, as reported by The Associated Press; “Please bear with your pharmacy teams today, where waits may be longer than normal and there may be delays to some prescriptions,” a spokesperson said.
Additional operational impacts were reported by government ministries or agencies in Singapore, Japan, Greece, and Norway, alongside disruptions to online bond payments in Colorado’s Arapahoe County and business communications in Hong Kong.
As recovery efforts progressed into Friday, The New York Times said that Microsoft issued apologies and explained that services were restored gradually to validate stability at each stage. Yahoo News reported that Microsoft confirmed in a management center announcement that the problem originated from an earlier network capacity upgrade that surfaced during a recent network change, and promised a complete report once an internal review is finished.

Following the restoration, advisory firm Infotech advised chief information officers to verify the recovery of business-critical workflows, integrations, and security controls before declaring normal operations restored, according to The Cullman Times.
Kate Wood, AVP of Research Development at Infotech, stated, “CIOs need a clear basis for telling the business it can return to normal operations. That means knowing which critical processes have been tested, where security coverage still needs checking, and which teams must continue using workarounds.”
Affected users seeking compensation may be eligible to apply for service fee credits under Microsoft’s service level agreement, depending on the product used and the degree of disruption, according to Yahoo News. However, claimants must actively contact support and provide relevant evidence for review.