A significant outage affecting core Microsoft 365 services, including Outlook, Teams, and Copilot, extended into its second day on Friday after initial repair efforts failed to restore functionality overnight. The disruption left institutions and individuals across the globe without essential communication tools, causing widespread delays in schools, courts, government agencies, and corporate operations. While Microsoft indicated on Friday morning that service availability was improving, the company previously stated that its initial fixes were “not taking as expected.”
Microsoft has not disclosed the specific root cause of the incident but indicated that the failure resulted from cascading server issues in multiple data centers. The company explained that unexpected unavailability of certain servers shifted excessive load onto remaining infrastructure, rather than a spike in user traffic. Microsoft officials tested configuration changes to address servers failing automated health checks and emphasized there was no evidence of a cyberattack being responsible for the downtime.
The incident has been compared by industry analysts to the major Facebook outage of 2021. Doug Madory, director of internet analysis at Kentik, noted that the recovery process has been uneven. The disruption has affected numerous sectors globally; in Australia, medical scans and exam schedules were cancelled, while in Japan, government staff were forced to rely on personal chat tools. In the UK, police forces could not process online crime reports, and parts of the National Health Service faced backlogs in patient referrals. Similar issues were reported in Germany, the Netherlands, France, Canada, and the United States, where services such as Via Rail and local law enforcement agencies experienced significant operational hurdles.
Experts suggest that recent workforce reductions and the increasing integration of AI in Microsoft’s codebase may have contributed to the difficulty in resolving the issue. Jake Williams, a faculty member at IANS, noted that the loss of institutional knowledge due to layoffs could be impeding resolution efforts. Microsoft’s annual filings showed a decrease in full-time employees from 228,000 in mid-2025 to 223,000 by June 2026, with an additional 9,000 jobs cut in July. Furthermore, CEO Satya Nadella recently stated that up to 30 percent of the company’s code is now generated by AI, raising concerns about the reliability and maintainability of these automated systems.
The outage has intensified debates among policymakers about the resilience of digital infrastructure. Chi Onwurah, chair of the UK parliament’s Science and Technology Committee, stated that the event highlights the dangers of dependency on US-based big tech firms for public services. While European countries are developing local alternatives to major productivity software, experts advise that organizations should focus on identifying mission-critical services and establishing independent fallback systems rather than abandoning large technology providers entirely.
The comparison to Facebook is apt, but this affected medical schedules. Real people missed appointments, not just social media.
In my country, the government had to switch to WhatsApp for official communications. It is not exactly secure, is it?
Another day, another cloud outage. These single points of failure need to stop being so central to everything.
The AI code generation comment scares me. How can you trust automated systems to run critical infrastructure without human oversight?
Does this mean we should all switch to local servers now? Feels like we are stuck in 1995, but maybe safer?
My entire office was offline for two days. We literally printed emails on paper. It was both funny and terrifying.