Photo: Pexels
“Microsoft 365 outlook down” became one of Indonesia’s top searches this week, and Indonesian users were far from alone. On Thursday afternoon US Eastern time (late Thursday night and Friday in Jakarta), Microsoft’s email and collaboration services began failing. Within an hour of Microsoft acknowledging the problem, Downdetector had logged more than 100,000 reports. The disruption lasted into a second working day.
What happened
Microsoft initially described the problem as affecting users in the Americas. It later said it “may affect Microsoft 365 service connectivity for users in any region”. Outlook mailboxes would not load, emails were delayed for hours, and Teams was disrupted until Microsoft said “mitigative actions” had restored it. Businesses, schools, hospitals and governments were affected. In New York, city emergency management warned residents that emails to city agencies could be delayed.
Microsoft paused all deployments, moved traffic off the affected infrastructure and restarted servers. By Friday morning it reported recovery in North America and Europe and “encouraging recovery signals in Asia Pacific”, with email queues “approaching normal levels”.
On Friday it gave the cause: “This was a technical issue not related to cyber or AI/agents. The incident was introduced by an issue during an earlier network capacity upgrade that surfaced following a recent network change.” Charles Lamanna, an executive vice president at Microsoft, apologised and said teams were reviewing “how we can prevent similar issues going forward”.
My view: these outages will keep happening
The explanation is reassuring in one sense: there is no sign of an attack, and no data breach has been reported. It is worrying in another. A routine change to Microsoft’s network was enough to disrupt work for organisations around the world for two days.
This is the trade-off of cloud services. A company that moves its email to Microsoft 365 or Google Workspace gets better security, lower costs and fewer IT headaches. It also loses control. When the provider fails, every customer is affected at once, and none of them can fix it. The same pattern has repeated in recent years with CrowdStrike, AWS and Cloudflare.
In Indonesia, government agencies, universities and banks have moved quickly onto a small number of global platforms. Few have tested what they would do if those platforms were unavailable for a day. Some practical steps:
- Have a backup communication channel. Know in advance how staff will communicate if email and Teams fail, whether by phone, an alternative messaging app or SMS.
- Keep offline copies of critical documents. Contracts, contact lists and procedures should be available locally.
- Ask your provider about service guarantees. Many organisations accept standard terms without checking what compensation is available after an outage.
- Practise. An outage drill once a year costs very little compared with a day of lost work.
Microsoft will publish a detailed review of what went wrong. Customers should review their own preparedness too.
Sources: ABC News; CNN; Tom’s Guide.
