Some achievements shouldn't have to be accomplished twice.
Eliminating sustained measles transmission in the U.S. ought to be one of them.
Yet here we are. More than a quarter-century after achieving that milestone, the country has lost its measles-elimination status.
The U.S. measles outbreak began on Jan. 20, 2025. Nearly 22 months later, on Nov. 17, 2026, health officials confirmed the loss of the elimination status the country had maintained since 2000.
We knew how to prevent this. We still do.
Measles didn't come back because the vaccine stopped working. We lost ground because fewer people were protected. IT has its own ways of making that mistake.
Nobody gets a standing ovation because the network stayed up, the backup worked or a ransomware attack never happened.
And therein lies the problem.
When prevention works, its value becomes almost invisible. Eventually, someone asks whether all those precautions are necessary. Could we spend less? Move faster? Let AI handle it?
In 2024, attackers got into Change Healthcare through a remote-access portal that lacked multifactor authentication (MFA). They deployed ransomware, disrupting prescriptions, medical claims and payments nationwide. An MFA check could have blocked that particular route into the system.
That's an awfully consequential place to skip a basic security precaution.
Microsoft 365 ran into a different problem in November. Testing hadn't caught a software defect because the test environment didn't match production conditions. When traffic shifted to gateways running with reduced capacity, those systems became overwhelmed. The protections intended to contain the trouble failed, too, and the outage spread to customers.
It's a reminder to test the ugly scenarios: reduced capacity, unexpected traffic shifts and safeguards that buckle under pressure.
Then there's UniSuper, the Australian financial institution whose Google Cloud subscription was inadvertently deleted in 2024. UniSuper had backups, but restoring data wasn't enough. Its cloud environment and configuration also needed rebuilding. Restoring full service took nearly two weeks.
The lesson? Test whether you can restore the operation, not merely retrieve the files.
Three incidents, all pointing back to familiar IT fundamentals: protecting access, testing under real conditions and making sure you can actually recover.
Public-health and enterprise IT data illustrate a common challenge: Effective safeguards must be maintained, tested and used.
Public health | Measles prevention
95.2% — U.S. kindergartners with recommended measles, mumps and rubella (MMR) vaccination coverage in the 2019–20 school year.
92.4% — MMR coverage in 2025–26, below the approximately 95% community-immunity benchmark. The decline left an estimated 280,000 kindergartners without documentation of completing the recommended MMR series.
Enterprise IT | Cyber recovery
78% — Surveyed IT and security decision-makers who said their organizations' cyber recovery plans focused more on restoring systems than maintaining critical business operations.
22% — Respondents whose organizations had formally documented and tested a minimum viable company — the essential operations needed to keep serving customers during recovery.
Sources: CDC, SchoolVaxView (2019–20 and 2025–26); Cohesity, 2026 Global Cyber Resilience Report.
But there's a better example worth celebrating.
After receiving a credible federal security warning in September, Kiteworks recommended that customers shut down production systems and took its own hosted services offline. The company subsequently discovered and patched a critical vulnerability, reporting no evidence of compromise.
I'll give Kiteworks credit. Taking systems offline before you know whether attackers have gotten in isn't an easy call, especially when customers are counting on those services.
That's thoughtful leadership. Not every warning warrants a shutdown, but sometimes prevention's inconvenience is a bargain compared with the consequences of doing nothing.
And those consequences are growing.
These days, a company can find itself in trouble because something broke at a provider three steps removed from its own operations. And finding a quiet time to take anything offline? Good luck.
That doesn't necessarily mean technology fails more often. But failures can travel farther, faster and affect more people.
When prevention works, its value becomes almost invisible. Eventually, someone asks whether all those precautions are necessary.

Of course, technology changes. Sometimes reinvention is essential.
There's a difference between improving safeguards and assuming you know better than the evidence behind them. Hubris creeps in when you're convinced you've anticipated everything that could go wrong.
Bob Dylan put it memorably in 1965: "He not busy being born is busy dying."
Fair enough. But moving forward doesn't mean abandoning everything that got you this far.
We don't improve public health by abandoning effective vaccination programs for theories unsupported by evidence. Nor do we improve IT by discarding the engineering principles and hard-earned experience that made our systems reliable.
The trick is knowing what needs to change — and what still works for a reason.
So, CIOs, here's a question: Which safeguards in your organization have become so successful, routine or invisible that people have stopped appreciating why they're there?
And when did you last check whether they still work?
James Alan Miller is a veteran technology editor and writer and Lead Editor for CIO News at Informa TechTarget. He directs coverage of enterprise technology strategy, AI, software, data, infrastructure and the business of IT.