Major global disruption in Outlook and Teams – problems could last a long time, according to experts

A large-scale disruption of the software company Microsoft's central cloud offerings is causing problems for many entrepreneurs and consumers. The connection problems affected users of Microsoft 365 services in all regions of the world, the US software company announced. Initially, Microsoft only investigated reports of problems from America. According to Microsoft, there is no evidence of a cyber attack.

The website Downdetector reported more than 130,000 outages in the US at its peak before problems spread to the Asia-Pacific region on Friday. A large rash of reported disruptions for Outlook was registered on the German platform “alletreibung” on Friday morning.

Not only the Outlook email service is affected, but also general services from the Microsoft 365 cloud package. In the Office communications service Teams, for example, the planning of meetings and calendar connections did not work properly. In addition, the AI ​​assistant Copilot, the management portal for administrators and interfaces to third-party providers - including AI services such as OpenAI and Anthropic - were temporarily disrupted.

According to recent information from Microsoft engineers, a process on the affected servers that is responsible for establishing secure connections was overloaded - as a result of which the computers no longer processed requests. The company also reversed the most recent configuration change, which coincided with the start of the disruption.

In the meantime, Microsoft seemed to be able to report success: some of the servers were being brought back online - “and they remain stable so far”. However, a setback followed shortly afterwards: servers that had been restarted threatened to fail again as soon as data traffic hit them again. A later attempt to start up parts of the infrastructure in isolation initially only produced partially positive results; According to the company, further countermeasures did not work as hoped.

Microsoft said it was looking for alternative ways to increase capacity given persistent imbalances in traffic control. It is unclear when all Microsoft 365 services will run without restrictions again.

IT security expert Dennis Kenji-Kipker said on ZDF that “tens of millions of users” were affected worldwide.

There was already a “hiccup” in Outlook on Thursday evening. Microsoft tried to fix this overnight, but it didn't work. “The cause is ultimately an internal network change,” Kipker explained. “That means some Microsoft servers were incorrectly logged off as being faulty.” The remaining servers then had to handle all global data traffic alone.

Microsoft must reverse this incorrect network change. In addition, the millions of requests would have to be blocked in order to be able to restart the servers. Microsoft will probably gradually increase the computing capacity in the data centers again - and as soon as enough servers are running stably, it will allow global data traffic again gradually and in a controlled manner. “We are talking about a major, widespread disruption here, which means that we will definitely have to live with the fact that basic communication tools are not working for a few more hours today.”

Tech journalist Chris Stokel-Walker told the British broadcaster Sky News that he would also expect problems at the weekend. A first indication is the fact that Microsoft has extended the timeout period for emails to 48 hours as a precautionary measure. This is the amount of time the system continues trying to send an email before giving up trying. “That 48 hours is probably an indication of how long they think it might take for the system to process those emails,” Stokel-Walker said. Some emails may never reach their destination.

Sensitive government areas are also affected by the disruption. In Australia, the secret service and the federal police reported failures. The Canberra government activated an agreement with Microsoft to ensure priority system restoration. Australian Home Affairs and Cybersecurity Minister Tony Burke said Microsoft was not expecting a cyberattack. Rather, it is apparently a faulty update, similar to the global IT breakdown caused by the security company CrowdStrike in 2024. The US House of Representatives also reported restrictions for its users.

The effects also spread widely throughout the economy. Artificial intelligence (AI) developers OpenAI and Anthropic investigated issues with interfaces to Microsoft services not working as expected. The US electric truck manufacturer Workhorse Group postponed the presentation of its quarterly figures due to IT problems. Microsoft previously stated that a number of servers were unexpectedly unavailable. However, this is not due to a sudden increase in data traffic. Just last month, a failure of the Azure cloud platform caused problems worldwide.