November 17, 2026

Microsoft announced the root causes behind the widespread outage that occurred in Microsoft 365 services last week, which led to disruptions that disrupted the interests of hundreds of millions of individuals and companies around the world.
“People” magazine quoted Charles Lamanna, Executive Vice President of the sectors responsible for “Copilot” and the agents and platform in the company, that the malfunction that occurred on Thursday, November 12, resulted from the intersection of three main factors combined.
Lamanna explained that the first factor is a technical flaw hidden within the process, which leads – under certain conditions of pressure – to the servers crashing and stopping responding. He explained that this defect is the result of a fix prepared specifically for the network portal servers in “Microsoft 365”, stressing that this defect and the fix are not related to the wider deployment operations of the “Windows” operating system. He added that the update related to this fix was installed last July, and underwent a secure deployment process that took weeks without detecting any problems, and it worked globally for more than three months normally.
The second factor was the sudden increase in the volume of loads on the network gateway on which Microsoft 365 services depend. Lamanna pointed out that the company relies on two independent and different network gateway systems to ensure isolation and high availability, which allows services to continue operating even in the event of faults.
Regarding the third factor, the Microsoft official explained that it was a decrease in server capacity, as about 14% of the portal’s server capacity was unavailable as a result of a separate problem that occurred the day before the incidents.
Lamanna confirmed that the coincidence of the sudden increase in loads with the decrease in server capacity led to the activation of the technical defect that had been hidden since July, causing a massive accumulation of requests until the servers stopped responding completely coinciding with the peak of traffic in the United States.
According to what People magazine reported, the problem began before 11:17 a.m. EST on Thursday, and the DownDetector website recorded about 20,000 reports of disruption of five of the company’s main services per minute, about two hours after the outage began.
Lamanna concluded by stressing that the company will issue a comprehensive final post-accident review later this week, to provide more details and transparency regarding the reasons that led to this defect.