What happened

Microsoft is trying to restore services after a global outage that reportedly disrupted access for tens of thousands of users S1. The problems hit Outlook email, the Teams messaging and meetings app, and Copilot, Microsoft's artificial intelligence software S1S2.

The company described the trouble as "connectivity problems" S1. In a series of updates posted on X, it said it was beginning to see "positive results" as it worked on a fix S1.

Al Jazeera reported that Microsoft said its servers went offline unexpectedly, disrupting Outlook, Teams, Copilot and other services S2. Neither report gives a country-by-country breakdown of who was affected, and no official figure for the number of disrupted accounts has been published S1S2.

What the affected tools do

Outlook handles email and calendars. Teams carries workplace chat, calls and meetings. Copilot is Microsoft's artificial intelligence software, and it was among the services caught up in the disruption S1.

That mix explains why a single connectivity failure is felt well beyond one application. A team that cannot reach its inbox usually cannot hold its scheduled call either, so work that depends on both simply stalls. For organisations that run their day on these products, the immediate cost is measured in lost hours rather than a line on a bill.

The AI dimension

Copilot's inclusion is the detail worth pausing on S1. AI assistants of this kind are not software sitting on a laptop; they run as online services, which means they are exposed to the same server and connectivity failures as email and chat.

That matters as more companies fold AI tools into daily workflows. The promise is speed on drafting, summarising and search. The trade-off is that the assistant goes down with the infrastructure behind it, at the same moment the rest of the working day does.

The money question

Neither report puts a financial figure on the disruption S1S2. There is no stated estimate of lost business hours, no confirmed count of corporate customers affected and no indication of whether any organisation has sought compensation or moved work to rival tools S1S2.

For buyers of cloud software, the practical questions are contractual rather than technical: how much downtime a service agreement permits, and what credit, if any, follows a breach. Those terms are set deal by deal and are not addressed in either report S1S2.

What to watch

The next useful signals are a confirmed cause and a firm user count. The "tens of thousands" figure comes from the reports rather than from a company tally S1S2. Watch Microsoft's X updates for confirmation that the fix holds, and for whether the disruption lasted through a full working day in the regions hit hardest.

Because the same products are used by businesses of very different sizes, the follow-up question is how quickly normal service resumes and whether Microsoft gives customers a detailed post-incident explanation.