Microsoft on Monday blamed a costly outage of its 365 services — including Outlook, Teams, SharePoint, Copilot and Defender — on complications related to a software defect introduced in July.
The technology giant said three factors combined to trigger the crash. It began around 12:30 pm Thursday with Outlook email accounts freezing, and caused millions of dollars in damage to organizations before ending Saturday morning.
“The update went through safe deployment and saturated globally over several weeks, during which time no issues were observed,” company executive Charles Lamanna said in a Monday social media post. “For more than three months, the change was deployed worldwide without impact.”
People were interested in these podcasts
Mr. Lamanna said the “latent defect” made the software system unable to handle certain loads of activity.
The other two factors driving the crash were reduced server capacity and Teams traffic shifting onto the same gateway shared with Exchange Online.
The explanation expanded on an initial theory the company released last week, tying the outage to an earlier network capacity upgrade that surfaced after a recent network change.
Multiple media reports found that the outage paralyzed or slowed courts, schools, airlines, brokerages, government offices and healthcare companies around the world that rely on the software.
Many switched to text messages, phone calls and other technologies to continue operating during the outage.
More than 430 million people use Microsoft 365 worldwide.
The Microsoft crash inspired a slew of online conspiracy theories. Microsoft addressed these rumors in a social media statement on Friday, calling the outage “a technical issue not related to cyber or AI / agents.”
There is no definitive estimate of damage costs caused by the outage, as companies continued to pay idle staff on top of recovery and rework expenses.
Anderson Economic Group has projected that lost productivity could approach $7 billion, a preliminary figure.
CyberCube, which provides online risk analysis for the insurance industry, estimated Monday that final global insured cyber losses will likely stay under $100 million. But the company set a cautious upper limit of $350 million.
CyberCube has dubbed the event “M364.” It said factors in its estimate included the outage’s relatively short duration, the absence of any reported data corruption and how quickly companies found alternatives to email for communications.
“This event once again reminds us how the technology world is concentrated in the hands of a few key providers,” Jon Laux, vice president of products and analytics, said in a statement. “But it also reminds us that not all layers of the technology stack are equally critical to business operations.”

Please read our comment policy before commenting.