What Caused the Microsoft Copilot Outage on June 1, 2026?

Microsoft publicly acknowledged the incident at approximately 11:27 a.m. ET, stating it was investigating "Microsoft 365 Copilot app load and timeout errors" . The outage lasted a total of 4 hours and 25 minutes, with service restored by early afternoon Eastern . During the disruption, users encountered app-load failures and timeouts, with some also seeing 503 errors indicating the service was temporarily unavailable .

Why Did Copilot Go Down?

Microsoft has not yet issued a definitive post-incident report confirming a single root cause. However, two contributing factors emerged during the outage:

• A routing failure directed Copilot traffic toward unhealthy infrastructure instead of healthy service capacity, causing access problems and timeouts .
• Lingering Azure infrastructure damage from severe thunderstorms that struck the West US 2 region on Friday, May 29, caused widespread power outages and disrupted Microsoft cloud services. Copilot was reportedly still recovering from that incident when the June 1 outage hit .

Microsoft confirmed it had received logs from impacted users and was reviewing them to identify the root cause . The company rerouted requests to healthy infrastructure as part of the mitigation .

A Separate MFA Outage on the Same Day

Compounding the disruption, a separate, unrelated outage struck Microsoft's Multi-Factor Authentication (MFA) setup and the My Sign-Ins portal (mysignins.microsoft.com) on the same day .

The MFA outage lasted roughly 3 hours and was resolved by 9:01 a.m. ET—before the Copilot incident peaked . Users attempting to set up MFA or access the My Sign-Ins page encountered 504 Gateway Timeout errors . Microsoft attributed this incident to cache failures in the underlying infrastructure and mitigated it by failing over to alternate healthy infrastructure . Tracking code MO1329260 was assigned to the incident .

A Pattern of Copilot Reliability Problems

The June 1 outage is not an isolated event. It extends a pattern of reliability issues that has dogged Microsoft's AI assistant:

• Over the last 90 days, Copilot logged 14 incidents—7 major outages and 7 minor—with a median duration of 1 hour and 18 minutes .
• On April 23, 2026, users experienced intermittent service issues for about 1.5 hours, with errors like "Having trouble with requests" .
• From April 28 to May 13, 2026, a 15-day minor incident prevented Copilot Chat users from accessing links to other apps .
• On February 25, 2026, scattered regional failures affected some users .
• In December 2025, regionally concentrated outages struck users in the UK, Europe, and Asia .
• In October 2025, a major Azure Front Door configuration error also took down Copilot services .

Microsoft has now suffered five major cloud outages in six months as of mid-April 2026, with Copilot frequently among the affected services .

Enterprise Risk: AI as a Productivity Dependency

The timing of the outage is significant. On May 30—just two days earlier—Microsoft launched a significant Copilot redesign focused on architectural improvements, promising 2x faster load times and better reliability in direct response to enterprise complaints about sluggish performance . The June 1 disruption immediately undercut that reliability narrative.

Microsoft is embedding Copilot deep into Word, Excel, Teams, Outlook, and other core tools, making AI a dependency layer in daily work. When that layer fails, users lose more than a chatbot—they lose AI functionality woven into the productivity stack .

The outage adds to mounting enterprise skepticism. Analysis suggests 40% of enterprises should defer Copilot deployment by 12–18 months due to reliability, security, and unclear ROI . Copilot's share of primary platform preference among U.S. paid subscribers reportedly fell from 18.8% in July 2025 to 11.5% by January 2026, driven by competitive pressure and product experience gaps . Meanwhile, Copilot faces intensifying competition from Google Gemini, OpenAI's ChatGPT Enterprise, and other embedded AI assistants, making each outage a credibility hit .