Microsoft has identified a latent software bug present since July as a root cause of the global outage that took down Outlook, Exchange and other Microsoft 365 services last week, according to a preliminary report the company issued on Monday and reported by
The Guardian (theguardian.com). The company said three factors combined to cause the disruption: the months-old bug, malfunctioning internal connections and reduced server capacity. Each might have been resolved on its own, but together they produced an outage that was harder to mitigate.
The disruption began on Thursday with US users reporting interrupted access to Outlook, then spread into a 24-hour blackout affecting millions of people worldwide. The Guardian reported that jails could not release prisoners, hospitals turned to paper records, schools lost contact with parents and police departments could not receive crime reports. Microsoft said it had restored service by Saturday and issued an apology, and did not attribute the blackout to a cybersecurity breach.
How three failures combined
In a post on X on Monday, Charles Lamanna, Microsoft's executive vice-president of Copilot, Agents and Platform, said a latent defect from July caused the company's servers to "become overwhelmed and unresponsive". Microsoft typically runs two networking gateways, the entry and exit points that connect separate systems. One was not working for Microsoft Teams, which overloaded the primary gateway, and more than 14% of server capacity was also unavailable because of an unrelated issue from the previous day.
"Taken together, the increased load and the reduced server capacity, caused the network gateway to encounter the latent bug from July," Lamanna wrote, according to The Guardian. "This caused requests to build up until servers ultimately stopped responding. This occurred during the peak window of traffic in the United States."
The Straits Times (straitstimes.com) adds detail on the sequence. It reported that the bug had been lurking undetected in Microsoft's data centres since July, and that routine rerouting of user traffic on 11 November caused the system to malfunction. Just after 8pm New York time, traffic from Teams was automatically rerouted to the separate system of servers that also handle Outlook traffic. Combined with the earlier loss of capacity, that triggered the bug, which prompted servers to exhaust their memory after an unexpected surge in traffic. By 11.30am on 12 November, as US business traffic reached its peak, Microsoft realised it had a problem.
Ars Technica (arstechnica.com) reported that the preliminary Post-Incident Report places the outage between 12 November at 16:20 UTC and 13 November at 16:10 UTC, just under 24 hours, with additional monitoring through 14 November. It said the gateway server fleet was operating 14 percent below optimal conditions, and that backup systems shifting requests away from unresponsive servers met the same lockouts on supposedly healthy gateways, leaving more than 80 percent of affected gateway servers frozen.
Why recovery took so long
Recovery was slower than usual because the defect predated recent changes. Lamanna said normal rollback, flighting and feature flags could not quickly revert a bug from several months earlier, and that automated recovery processes were not designed to restore service from such a broad unresponsive state. The Straits Times reported that Microsoft spent about 15 hours before it began rolling out a fix, and that at the nadir more than 80% of the servers behind Outlook were unresponsive. Ars Technica noted that restarting servers was ineffective because large request queues immediately triggered the same resource exhaustion.
The Straits Times also reported that the bug escaped testing because simulated traffic patterns and lab hardware did not reflect real-world conditions, and that the defect is more destructive on newer hardware with more processing power. Microsoft shares were down about 1% at 2.29pm on 16 November, the paper reported.
Commercial impact
The immediate cost falls on organisations that depend on Microsoft 365 for email, chat and file access. The Straits Times reported that work at many corporate and government clients ground to a halt, with inboxes failing to load, Teams invitations going missing and staff resorting to phone calls and pen and paper. The Guardian quoted Matt Hull, vice-president of cyber intelligence and response at NCC Group, who said: "This is a prime example of the importance of not only cyber resilience, but business resilience." He added that reports of hospitals, schools and other essential services being affected demonstrate the wide impact a system outage can have, regardless of the cause.
For operators, the episode is a reminder that concentration on a single productivity suite turns a vendor's internal defect into an operational event. The sources do not quantify financial losses, and no estimate should be inferred from them.
Risks and unknowns
Microsoft has said it will publish a final report, and Ars Technica reported that the full final incident report is expected on 20 November. That document may clarify why regional isolation systems appear to have failed, a question the current report does not address. Microsoft said it has added capacity and traffic-distribution guardrails and extended delivery timeouts on affected emails. Whether those measures address the underlying testing gap described by The Straits Times remains unconfirmed.
What happens next
The final report on 20 November is the next verifiable checkpoint. Until then, the preliminary account supports a narrow conclusion: a defect that sat undetected for months, a standard failover reroute and a capacity shortfall interacted in ways Microsoft's testing and automated recovery did not anticipate. The company has not attributed the outage to a cyber breach, and Ars Technica reported the incident's final status as "a technical issue not related to cyber or AI/agents".
Sources and References
Why It Matters
The outage shows how a single latent defect inside a widely used platform can cascade into operational disruption for hospitals, schools, police departments and corporate clients. For founders and operators, the event makes vendor concentration and failover planning a board-level question rather than an IT detail, because the failure originated months before it surfaced and could not be undone by rolling back recent changes.
FY Outlook
Microsoft has said it will publish a final incident report, which Ars Technica reported is expected on 20 November. That document may explain why regional isolation systems appear to have failed and set out corrective actions. Until it appears, the preliminary report is the most authoritative account, and the company's added capacity and traffic-distribution guardrails remain untested in public.