Charles Lamanna, Executive Vice President of Copilot, Agents, and Platform at Microsoft, summarized the root cause analysis in a post on X[ indicating that there were three issues in Microsoft's networking gateway servers that combined to cause the outage:
He stated that the automated rollback mechanisms couldn't mitigate issue (1) as the change causing the error was deployed in July 2026. As the code fix was deployed, traffic overwhelmed the servers coming up - an instance of cascading failure.
Microsoft 365 customers have been directed to MO1550821 in their admin center tool for further updates..
Microsoft said that the outage was not caused by a cyberattack or AI agents,[3] and blamed the outage on settings applied to servers after an update. They said that a process that establishes secure connections was becoming overwhelmed, which led servers to stop handling requests. They began rolling back the recent network configuration change.[4]