Microsoft's outage shows resilience is a business decision
A global Outlook, Teams and Copilot failure is a reminder that bundling mail, meetings and AI under one vendor is a bet, and the board should own it.
Photo: KK IN HK / Wikimedia Commons (Public domain)
What happens to your company when the email goes dark, the meeting invites vanish and the AI assistant goes quiet, all at once, from the same vendor? On Thursday, a lot of organizations found out.
Here is what we know. According to Hardwareluxx, Microsoft began investigating Exchange Online mailbox problems on Thursday, November 12, first in the Americas. The trouble then spread, and Downdetector logged reports from 74 countries. Outlook alone drew nearly 130,000 reports at the peak. Microsoft said a large number of servers failed unexpectedly, and the remaining infrastructure was overloaded. It also said servers brought back online fell over again once traffic returned. It has found no evidence of malicious activity. By Friday midday, Tagesspiegel reported signs of improvement and said core Teams and Copilot functions had been restored. There was still no official all-clear.
The cost isn't downtime, it's concentration
Outages happen. Hardwareluxx notes that Microsoft had a roughly ten-hour North American incident in January and an Exchange Online failure in late August. None of that is the interesting part. The interesting part is what a single failure took with it.
Email, calendars, meeting scheduling, the admin portal and the AI assistant were all hit by the same incident. Even the connectors that let OpenAI's and Anthropic's chatbots read mail and documents in Microsoft 365 were disrupted, according to t-online. That is the new wrinkle. In our view, the AI layer isn't a separate safety net. It sits on top of the same pipes.
We think this is the real lesson of the "copilot everywhere" pitch. Every feature you add under one roof makes the suite more convenient. It also makes the suite a bigger single point of failure. Convenience and fragility are the same product decision.
The people who felt it weren't browsing memes
The reported fallout wasn't trivial. Australia's intelligence service and federal police reported outages, and Canberra activated an agreement with Microsoft to get priority restoration. Per the Tagesspiegel, citing Bloomberg and Nikkei, traders in Asia switched to fax machines and phones to report trades. A US electric truck maker, Workhorse Group, postponed its quarterly results.
Fax machines are, frankly, the best resilience plan in this story, and nothing in the reports suggests it was a designed one. Our view is that this was luck and improvisation, not planning.
The most pointed reaction came from Paul Spain of Gorilla Technology. As t-online reports, he told the BBC that if no cyberattack was behind the outage, it would come down to incredible incompetence. We'd be more careful. Microsoft hasn't yet explained the root cause, and a recently deployed network configuration change that it rolled back is only a suspect. What we can say is that the repeated incidents this year make "it's rare" a hard argument to sustain.
What a sensible company does on Monday
We aren't telling anyone to rip out Microsoft 365. For most organizations that would be its own self-inflicted outage. The point is to treat resilience as a line in the budget, not a footnote in the IT runbook.
- Know your fallback for mail and meetings. If the answer is "wait," someone should have signed off on that consciously.
- Keep out-of-band contacts. If your incident chat lives in the thing that's down, it isn't an incident channel.
- Don't wire critical processes to an AI assistant that shares the suite's fate. Automation that stops when email stops isn't automation, it's a dependency.
- Prepare staff for scams. Banks, universities and other institutions warned that fraudsters pose as IT helpers during outages, according to t-online.
Whether the cause turns out to be a bad config, a cascading overload or something else, the board-level question is the same: how much of the business are we comfortable hanging on one vendor's worst day? Some will say that's an acceptable price for simplicity. That's a legitimate answer. It just shouldn't be an accidental one.
Our verdict: the outage isn't proof that Microsoft is uniquely unreliable. It's proof that suite consolidation turns reliability into a strategic risk, and the AI era makes the bundle tighter. Decide on purpose, write it down, and test the fallback before the next Thursday.
No topics yet: start the first one.
More stories
The DRAM collusion suit survived, and AI can't be the alibi
A judge refused to dismiss the case against Samsung, SK Hynix and Micron. The AI boom explains some of what happened to memory, but it shouldn't shield the oligopoly from scrutiny.
United's AI agent ban is about the customer, not safety
United says it's protecting travelers, but its CEO's "another middleman" line gives away the real fight: who owns the relationship. Expect licensing, not a permanent wall.
CoreWeave's 16% yield is the market sending AI a bill
When a neocloud's debt yields more than 16%, the AI boom stops being a story about demand and becomes one about who can afford to keep borrowing.
Nvidia's real risk isn't Huang's affair. It's having no plan B
Whatever the probe finds, a company this central to AI cannot run on one man's judgment. Its governance has to treat the CEO's exposure as a security and investor matter.



