It was a strange week for anyone who relies on technology. Microsoft 365, Workday, a major bank, an airline, Apple and the U.S. air traffic system all failed within six days, and officials say none of it was an attack. At the same time, two well-known security voices are warning that attackers are moving faster than ever while the firms that clean up after breaches are swamped. This week we look at what both mean for a small business, plus a KPMG Canada breach, a fresh warning about IT providers, and a stat worth remembering. Let's get into it.
🗞️ STORY OF THE WEEK
A Week of Outages Shows Why You Need a Plan B for Your Providers
Between November 12 and 17, a run of high-profile failures hit organizations most of us treat as always on. Microsoft 365 (Outlook, Teams, Exchange Online) was down for nearly 24 hours. Workday customers were offline for about 18 hours. The U.S. air traffic system halted flights nationwide, Amazon's main U.S. cloud region had a three-hour disruption, JPMorgan Chase delayed payments for around seven hours, and Apple and United Airlines each had outages of their own. The NSA's classified network was also degraded.
Officials say none of this was a cyberattack, and so far every company that has explained itself points to something internal. Microsoft's preliminary report blames a hidden software flaw from July that only caused trouble when a Teams failover sent extra traffic to servers that were already running at reduced capacity. Workday says a backup-site test accidentally connected to its live databases. The FAA points to a newly installed router. Some of these explanations are still preliminary, and Amazon has not given a cause. The Cloud Security Alliance notes that in several cases the backup and failover systems meant to protect people helped trigger the failure.
The lesson for a business owner is not about who to blame. It is that you cannot see inside your providers, and when they fail you are left waiting. Security journalist Nicole Perlroth, speaking about the week, called it an opening chapter and pointed to brittle infrastructure and rushed updates. Whether or not that is right, the practical point holds: outages and attacks look the same from your desk, because in both cases your tools stop and you do not control the fix. One more thing to watch: outage days are prime time for fake 'reset your password' emails.
What it matters
Most small businesses run email, files, payroll, banking and sign-in through a handful of outside providers. A failure at any one of them can stop your work, even if nothing is wrong with your own systems. The question is not whether it happens again, but whether you will know what to do when it does.
What can you do
This week, write a one-page 'bad day' sheet: your five most important outside services, the phone number or status page for each, who on your team decides what to do, and how you will reach customers and staff without email. Print it. Ask your payroll and banking providers what their plan is if they go offline for a day, and tell your team to treat any 'reset your password' message during an outage as suspicious.
⚡THE RIPPLE EFFECT - Three stories, and what they actually mean for your business
Attackers Are Getting Faster, and the Firms That Respond to Breaches Are Overloaded
Security consultant David Kennedy of TrustedSec has described a test in which a swarm of AI agents broke into a company in about 14 minutes, and the company did not notice for roughly four and a half hours. He says the systems most businesses run were never designed to handle that speed. We could not independently confirm that incident, so treat it as a warning, not a measurement. Separately, Perlroth says incident response firms, the specialists you call after a breach, now have backlogs of about six months. Her claim is her own and is not an independently measured industry figure, but it matches what many firms are saying about demand.
What it means for you: If you are hit by ransomware tomorrow, the expert you call may not be able to start tomorrow. Find out now who your plan is, what they can promise, and whether your cyber insurer has a panel of responders you can use.
KPMG Canada Says Client Tax Documents Were Accessed in a Breach
KPMG Canada confirmed on November 18 that some client tax documents were accessed in a cybersecurity breach and that it is contacting affected clients. It has not said how the intrusion happened or how many clients are affected, though it says it is confident the data is not online or being misused. It follows an October alert from the Canadian Centre for Cyber Security about fake IT-support calls that steer staff to spoofed sign-in pages.
What it means for you: If you hold client files, such as an accounting, legal or advisory practice, make sure your help desk or IT provider verifies who is calling before resetting a password, and that you could quickly tell which client files were touched.
A New Warning About IT Providers: Patching Alone May Not Be Enough
Following last week's ScreenConnect story, security firm Huntback reports that a flaw in N-central, a tool many IT providers use to manage customers' computers, was being exploited within hours of its details going public on November 12. It found 21 live servers secretly running cryptocurrency mining software, four of them in Canada. Nineteen of the 21 had already installed the latest fix, so patching did not undo a compromise that had already happened.
What it means for you: Ask your IT provider whether they use N-central, whether the older feature at the center of this flaw is switched off, and whether they have checked their systems for compromise, not just installed the patch.
6 months
That is how long Nicole Perlroth says many incident response firms now have a backlog for new clients, as AI speeds up the discovery and exploitation of weaknesses. If true even in part, a retainer or a pre-arranged agreement matters more than a phone number in a search engine. Ask your provider, your insurer, or both, what response time they can actually commit to.
Source: Nicole Perlroth’s statements on X and in interviews, as reported by CNN and JournalArta, November 18, 2026. This is her own account and not an independent industry measurement.
✅ ONE THING TO DO THIS WEEK
Line up your Breach Help Before You Need It
Most small businesses have no idea who they would call if their systems were locked or data was stolen. This week showed two reasons that is risky. Outages and attacks can hit with no warning, and the specialists who respond are reportedly very busy.
This week, decide on three things. First, who is your first call (your IT provider, your cyber insurer's hotline, or a response firm), and what do they promise on timing? Second, are your backups stored somewhere an attacker cannot reach, and have you tried restoring one? Third, who on your team makes the call to shut things down, notify customers or contact the privacy commissioner? Put the answers on the same one-page sheet as your outage plan.
The best time to find an incident response team is before you need one.
📌 QUICK BITES
Norwegian energy company Equinor is containing a cyberattack on its internal IT systems. Oil and gas production was not affected because operations are kept separate from office systems, though staff face disruption to email and business apps. If email and your business apps went down for a week, what could still run?
The U.S. cyber agency CISA says four actively exploited flaws include one in LiteLLM, a tool for managing access to AI models. Attackers are also hijacking AI model access to run up bills. If you or a vendor self-host AI tools, add them to your list of things to patch, and check for Synology storage devices, another product on the list.
California has proposed requiring serious AI incidents to be reported within 48 hours, along with a state power to order an emergency shutdown. The rules target the biggest AI developers, but expectations tend to flow down. Know how you would switch off any AI tool your business relies on.
JPMorgan Chase is also dealing with a breach at a third-party vendor that held marketing data. Your suppliers hold your customers' information too. Ask yours what they keep and how fast they would tell you.
💬 PARTING THOUGHT
A week like this makes it tempting to hunt for one cause, one villain, one fix. The more useful habit is quieter: assume something you rely on will fail, and decide now what you will do.
Some failures will be attacks and some will be a bad update on a Tuesday. Your customers will not care which. They will care whether you kept serving them.
Resilience is a plan you wrote before the bad day.
See you next Friday and stay
TryberResilient
1