Microsoft Copilot suffered a possible outage on Thursday, June 11, 2026, after more than 4,500 users reported problems on Down Detector around 2:03 p.m. PT. The surge pointed to a real disruption rather than a brief blip, with many users saying they could not get the service to load properly.
The issue landed in the middle of the afternoon, when people turning to Copilot for work or quick searches were most likely to notice it. Most reports to Down Detector centered on website problems, and the outage disrupted service for thousands before engineers stabilized the platform.
Down Detector tracks service problems by collecting status reports from multiple sources, so the spike in complaints gave an early signal that something was wrong. Microsoft later acknowledged the June 11 outage on its Service Health dashboard and said a broken software update deployment caused the problem, then rolled back the change to restore global access.
That account leaves one thing unresolved: Microsoft has not released an official post-incident review for the outage, even though it says Copilot is fully operational today with no widespread system outages. The pattern of 503 Service Unavailable errors, authentication loops and a sudden drop in user reports suggested the failure sat deep in backend infrastructure, and for anyone still seeing connection loops or failing prompts, the lingering problem may simply be cached data from Thursday's crash.
By Friday, June 12, 2026, the service was back to normal, but the size of the report spike shows how quickly Copilot users react when a core AI tool stumbles. For now, the outage is over; the explanation is partial, and the postmortem is still missing.
