Microsoft attributed the November 12–13, 2026, outage to a technical issue connected to an earlier network-capacity upgrade and a later network change. The disruption affected Outlook, Teams and Copilot; Microsoft said it was not caused by a cyberattack or AI agents. Recovery continued into Friday, while some businesses used phone calls and fax to keep work moving.
31
35
36
What Microsoft said went wrong
Microsoft initially reported server problems and said some servers had become unexpectedly unavailable. It later described a process on affected servers that establishes secure connections becoming overwhelmed, leaving servers unable to process requests.
9
37
47
The company said the incident had been introduced during an earlier network-capacity upgrade and surfaced after a subsequent network change. That explanation connects the outage to infrastructure changes, but the available reporting does not provide a complete technical post-incident account.
31
35
36
Microsoft ruled out a cyberattack and said the issue was not related to AI or AI agents. OpenAI and Anthropic reported problems with integrations that depend on Microsoft services; that does not mean their entire AI services were unavailable.
31
5
6
Which services were affected, and when did they recover?
Outlook, Teams and Copilot were among the services affected by connectivity problems, alongside other Microsoft 365 services. The disruption became evident in the United States on Thursday, November 12, with user reports of Outlook problems peaking around 1 p.m. Eastern.
3
5
7
The outage lasted more than 12 hours and extended into Asia-Pacific business hours on Friday. Microsoft reported improving availability and later described “broad recovery and stability,” but those updates do not establish the exact time every affected customer regained service.
1
18
31
How the outage disrupted trading and business
The outage made everyday communications harder for businesses that rely on Microsoft email and collaboration tools. Traders used phone calls, WhatsApp and fax to communicate; reporting also described businesses using Bluetooth as a workaround. Most organizations cited in Reuters reporting said their core systems remained operational, though some trading and administrative work slowed.
19
46
The disruption affected financial firms and exchanges as well as other businesses. Traders in Singapore reported that some banks could not complete financing deals, while Euronext, Nasdaq and Deutsche Börse reported email problems.
55
In Taiwan, Reuters reported that one financial firm used fax to confirm trades and another cancelled a bond repurchase transaction. The available reporting does not establish that the fax-confirmed trades were short-term bills.
18
The outage also affected corporate communications. Aviva said it had backup arrangements for earnings calls, including alternative phone numbers and Bluetooth transfers for recordings. Workhorse Group postponed a scheduled earnings release to November 16, citing business disruption.
20
30
The broader lesson: communications dependencies matter
The incident showed how a problem in widely used workplace services can ripple into trading, government communications and business administration, even when an organization’s core systems keep running. Reuters described the disruption as a reminder of global commerce and public services’ reliance on a small number of technology providers.
19
The reported problems with Microsoft integrations used by OpenAI and Anthropic illustrate a related dependency: connected tools can be affected when the underlying service they rely on has an outage, without those tools being the cause.
5
6
31