Microsoft said Friday it still has no timeline for a full recovery from a global Microsoft 365 outage that has stretched past 21 hours across 74 countries and territories, even after addressing the underlying technical fault. The company apologized to customers and partners and attributed the breakdown to a flaw introduced during an earlier network capacity upgrade that surfaced after a recent network change, ruling out a cyberattack, artificial intelligence or automated agents. While email queues are approaching normal levels as engineers restore service in stages to prevent servers from stalling again under returning traffic, Microsoft extended its email timeout window to 48 hours to keep delayed messages from bouncing.
The disruption, which began at 16:17 UTC Thursday, knocked out Exchange Online, Outlook, Teams, SharePoint, OneDrive, Copilot and Defender XDR, alongside Microsoft connectors in OpenAI and Anthropic tools. Combined problem reports across five Microsoft services reached 19,650 a minute on Downdetector on Thursday afternoon compared with fewer than 15 on an average Thursday, with Outlook complaints peaking above 140,000, or more than 2,000 times normal levels.
The outage drew scrutiny in Washington, where a US lawmaker asked Microsoft for information on the incident, Politico reported, while Brazil’s Supreme Court suspended legal deadlines nationwide at the request of the national bar association. Earlier, the failure forced Britain’s National Health Service to cancel some procedures and revert to paper records, UK pharmacies to warn of prescription delays, New York to pause Friday evictions, Euronext to tell listed companies to communicate by phone, Goldman Sachs employees to rely on backup email and Australia’s government to invoke a priority restoration agreement after its intelligence agency and federal police were affected.