It Wasn't Hackers or Rogue AI. It Was Something on the Risk Log.

The warning signs are often documented before a transformation goes wrong. The harder question is who owns the decision to proceed.

Everything Is Broken. Or Is It?

Last week it felt like the whole internet had a bad week at once. Microsoft 365 went down across more than 70 countries. Workday locked employees out in the middle of open enrollment. The FAA stopped every departure in the country. On Monday, Chase and United took their turns. And on Tuesday, even the NSA's classified internal network went down.

Naturally, the theories followed. It had to be a coordinated cyberattack. It was AI agents running wild. It was Friday the 13th. One AP story simply asked whether everything is broken.

Honestly, I get it. These stories are still unfolding, and some of what we think we know will change. But a villain is a strange kind of comfort. If someone did this on purpose, even someone frightening, then at least someone is in control. The alternative is harder to sit with: the systems that keep planes in the air and paychecks landing are complicated, a little fragile, and sometimes break by accident. Most of us would rather have an enemy than an accident.

The explanations so far are a lot less exciting. Microsoft, the FAA and the NSA have all said there's no sign of a cyberattack. What they described instead was the most ordinary thing in the world: a change that went wrong. A bug that had been sitting there for months. A network upgrade. At the NSA, scheduled maintenance to apply a software patch.

That ordinary part is the part worth talking about. It isn't really a story about other companies. It's a story about how most of us run our own transformations.

The One Where the Warnings Were in Writing

Of all last week's outages, the FAA's is the only one where we know the risk was written down ahead of time.

The ground stop started during routine work. A contractor was moving part of the FAA's telecommunications network from copper to fiber, and a newly installed router started flooding the system. Seventeen of the 22 centers that manage high-altitude traffic began to wobble. It became the longest nationwide ground stop since September 11.

None of that was a surprise in the abstract. Government watchdogs had been warning about this network for years, from an inspector general review of an outage during an earlier fiber upgrade in 2010 to a GAO report published two months ago. The GAO had already reported that 51 of the FAA's 138 air traffic control systems were unsustainable.

The investigation isn't finished, and I'm not going to guess at who did what wrong. But the risk of upgrading this network was known, documented and funded. The work went ahead anyway, and something broke in exactly the kind of place people had warned about.

You Have One of These Too

I don't know whether the other outages had warnings on record. I'd be surprised if nobody had a worry somewhere, but that's a guess, and it isn't the point.

The point is closer to home. If you've ever run a go-live, a cutover or a migration, you know where the soft spots are. They're in the risk log. They came up at the last steering committee. Someone on the team has been saying "I'm a little nervous about that weekend" for a month.

In my experience, most transformations aren't blindsided by risks nobody saw. They make the call with the risk in full view, and they go ahead anyway.

Today's Deadline Beats Next Month's Maybe

Why would smart people go ahead with a risk they can see? Some of it is the usual. We tell ourselves it won't happen to us, or that if it does, it won't be that bad.

But I think the bigger reason is simpler. The deadline is true today. The budget is true today. The date on the steering committee slide is true today. A risk is something that hasn't happened yet. Put something certain next to something hypothetical, and the certain thing wins almost every time.

So the pressure to get it done quietly becomes pressure to look past the risk. Nobody decides to ignore it. It just keeps losing to things that feel more real.

Nobody Throws a Party for the Outage That Didn't Happen

There's another problem. Risks don't make good stories.

"We hit the date" gets repeated at the town hall. "We spent two weeks reducing a risk that never showed up" doesn't. If you do the work and nothing happens, it can look like you wasted two weeks. If you skip it and nothing happens, you look decisive. The only time anyone sees the value of handling a risk is when somebody else didn't.

That's an incentive problem, not a character flaw. And it's why a known risk without a clear owner tends to stay right where it is, in the log, until the day it isn't.

Pick Your Risks on Purpose

The answer isn't to remove every risk. If that were the bar, no transformation would ever ship. The answer is to make a visible decision about each risk that matters: accept it, reduce it, or move the date. Whichever you pick, put a name next to it.

Before any switch on your plan, ask three questions:

  • Which known risk are we choosing to accept?
  • Who is accountable if it happens?
  • What will we watch in the first hour, and who is watching?

They're short enough to ask in a steering committee and specific enough to change a decision. Most of the time the answer will still be to go ahead, and that's fine. The difference is that you went ahead on purpose.

Before 2027 Locks In

You'll never hear about the risks that teams handled well, because nothing happened. Last week we all got a very public look at the other outcome.

Most of us are locking 2027 plans right now. A lot of organizations also freeze changes over the holidays, which tends to push the riskiest switches into January, right when every new initiative is starting. So before your plan locks, find the risks you're planning to live with and put a name next to each one.

It's a less exciting story than hackers or rogue AI. It's also the one you can actually do something about.

Written by Scott Sager

Founder of ChangeMap, after nearly two decades leading strategic initiatives and organizational transformations across public and private sectors.

Ready to see where your judgment counts most?

Start with what your transformation already produces. See the read, the reasoning, and the moves that matter.

Start Free

Keep reading

All posts →