Microsoft's Email Broke. America Asked for the Rest of the Day Off.

America's work email stopped working in the middle of Thursday. Exchange Online — the Microsoft system that runs Outlook for businesses, schools and government offices — started failing at 11:17 a.m. Eastern (8:17 a.m. Pacific), and the reports came in like weather. By 1:27 p.m. Eastern, Downdetector had logged more than 130,000 Outlook problem reports: one of the biggest email disruptions of the year, arriving at the exact hour the workday runs on email.

Every door into the mailbox was hit — Outlook on the web, the desktop app, phone sync, all of it named in Microsoft's own notice. Reports began spiking just after noon Eastern, passed 83,000 before 1 p.m., and kept climbing. Microsoft 365 reports peaked above 47,000. Copilot drew thousands of its own. Even the admin center, where IT departments were sent to follow the incident, was reported unreachable at times.

Microsoft's word: servers, not traffic

Microsoft confirmed the outage on its @MSFT365Status account: incident EX1550543 for the mailbox failure, and a broader advisory, MO1550821, for Microsoft 365 connectivity overall. The first post said the company was investigating issues accessing Exchange Online mailboxes "through multiple connection methods in the Americas." Inside the hour, the map grew: the issue, Microsoft said, "may affect Microsoft 365 service connectivity for users in any region." Canada and the UK were reporting it too.

The working theory moved as the afternoon went on. First, Microsoft said a portion of its infrastructure was experiencing "elevated load" across multiple regions. Then came the refinement: the load, the company determined, "isn't driven by an increase in traffic, but by a number of servers being unexpectedly unavailable" — the machines left standing were straining under the load the missing ones should have carried.

The response, per Microsoft's updates: all deployments and configuration changes paused while engineers review recent changes for any connection to the failure, connection traffic redistributed across its worldwide infrastructure, and affected servers restarted to bring them back into service. The wider advisory's list of potentially impacted services runs past email — SharePoint Online, Microsoft Copilot and Defender XDR are on it — and the damage jumped Microsoft's fence entirely: Reuters reported the disruption also hit connectors used by OpenAI's and Anthropic's AI services.

Recovery underway: reports are way down, no cause yet

As of Thursday evening, the incident is still open — but it is bending the right way. In its latest update, at 5:59 p.m. Pacific (8:59 p.m. Eastern), Microsoft said it was seeing positive results restarting the affected infrastructure in isolation and was monitoring for stability. The report curve agrees: Downdetector put Outlook reports at about 28,000 around 9 p.m. Eastern (6 p.m. Pacific), down from the 130,000-plus peak, in USA TODAY's live count. The recovery is partial — Microsoft says users may still hit disruptions as servers come back — and there is still no root cause and no restoration time. There is nothing to fix on the user's side of the screen; the failure is inside Microsoft's infrastructure. This story will be updated in place when Microsoft closes EX1550543 or posts a cause.

Teams survived. Nobody was happy about it.

The outage was the news. The reaction was the story. "Outlook is down" climbed to the top of Threads' tech panel on Thursday — not with troubleshooting tips, but with a demand: "If Outlook is going to go down, it should take Teams with it." A second post got straight to the point: "Since Outlook is out Nation wide, I think everybody should go home early from work today."

Girlboss, with 2.1 million Instagram followers, ran a three-slide carousel that said the quiet part in order: "Outlook is down but not Teams." "Whoever is working at Microsoft rn please just shut the whole thing down for the day." "Manifesting Teams goes down with Outlook today." The Greensboro Police Department joked on Facebook that with the email system down, the workday ends early — and commenters from other states and Canada treated a police department's post as official permission to leave. Aaron Parnas's breaking-news explainer drew 34,000 likes, its comments full of workers confirming their own inboxes were dead, and celebrating it.

One correction to the joke, because it matters to the record: Teams surviving is broadly true, not absolute. Teams drew a few thousand Downdetector reports of its own on Thursday, and Microsoft had already spent its morning on a separate North America Teams disruption, resolved around midday — roughly forty minutes before the email failure began. Thursday was a two-outage day at Microsoft. Email was just the one America wanted to keep.