Connectivity problems began shortly before 1 p.m. ET on Thursday and left millions of people without access to Outlook email, Teams meeting tools and other staples of modern office life.

According to Downdetector, outage reports reached 74 countries and territories in about an hour. Microsoft said the outage was not caused by a cybersecurity breach, with executive Charles Lamanna stating there was no indication it resulted from malicious or cyber activity.

CNBC reports the disruption lasted more than 20 hours, which it calls one of Microsoft's biggest-ever server issues.

The outage laid bare how much of daily work, government and civic life runs on a single company's infrastructure. Microsoft said in January that it had 450 million paid subscriptions to its Microsoft 365 suite, which was the worst affected.

Banks, governments, health services, railways and courts all reported disruptions across Microsoft products.

The New York Times notes that parents struggled to contact their children's schools, some government services faltered and many office workers could not access email as the disruptions stretched into a second day.

Consultancy Control Risks' James Owen told CNBC this highlights a business resilience and dependency issue that has been decades in the making.

The evidence of scale came from official channels around the world.

Several UK police forces, including Merseyside, Greater Manchester, Bedfordshire, Dorset and East Hertfordshire, told the public to report issues by telephone rather than online forms, urging callers to dial 101 for time-critical matters.

The New York Times reports the UK's NHS said the outage had a disruptive impact, Australia's intelligence and police forces said operations were affected, and the Supreme Court of the Philippines and the European Court of Human Rights warned of possible delays.

CNBC reports Workhorse Group and Richemont delayed earnings events, while Air India temporarily rerouted customer emails.

The outage reached beyond email into the AI layer of everyday work.

OpenAI said it was facing issues with Microsoft connectors, with tools including ChatGPT not working as expected, and thousands of Claude users reported issues Friday morning before Anthropic said a fix had been deployed, according to CNN.

Microsoft reported in 2025 that Copilot had more than 100 million monthly active users, and Copilot was among the affected services.

CNBC also notes an Australian minister compared the incident to the CrowdStrike update failure that caused one of the world's biggest IT outages in 2024, and that Microsoft shares dipped about 0.8% in premarket trade.

Microsoft says it is seeing encouraging recovery progress, warning that some users may still experience intermittent issues.

CNN reports email delivery performance is improving across a growing portion of affected infrastructure, with broader improvements in North America, sustained improvements across much of Europe and encouraging recovery signals in Asia Pacific. CNBC reports that by around 9 a.m.

ET Friday email queues were significantly reduced and approaching normal levels, with the email timeout window extended to 48 hours to protect delayed messages. The New York Times reported by 9 a.m.

Eastern that service was being restored and in some cases approaching normal levels.

Microsoft has not given a timeline for full restoration, and per the Times of India its service health remains variable, with Outlook, Teams and other Microsoft 365 apps working for some users while others still stumble.

The Times of India reports a patch to a component handling authentication and routing showed positive results, and that some servers unexpectedly went offline, straining remaining infrastructure; servers that were revived went down again once traffic returned.

Microsoft has promised a Post-Incident Report with the root cause once services are restored, and IT admins can track the incident under MO1550821 in the Microsoft 365 admin center.