DHS Chief Says AI May Be Behind Wave of US Outages — Or Not
DHS Secretary Markwayne Mullin says AI may link the FAA grounding and outages at JPMorgan, Microsoft and United — but admits he doesn't know, citing 'one guy' as his source.
By Nathan Brooks
3 min read

What's News
- DHS Secretary Markwayne Mullin said on Fox News Monday that new AI software might explain the November 13 FAA grounding and outages at JPMorgan, Microsoft and United.
- Mullin cited 'one guy that's very smart with tech' as his source and admitted he doesn't know for certain.
- The FAA halted all US air travel for several hours on November 13.
- Mullin said he is '98%' confident the US is not under a cyber attack.
- DHS has more than 200,000 employees and an annual budget over $100 billion.
The US Department of Homeland Security says new AI software might have caused the nationwide flight grounding last Friday — and the outages at JPMorgan, Microsoft and United. Or it might not. DHS Secretary Markwayne Mullin, speaking on Fox News Monday, admitted he really doesn't know.
His sourcing? "One guy that's very smart with tech" — not Elon Musk, Mullin specified — told him AI integration into IT stacks could be the common thread behind the recent failures. And "that kinda makes sense to me," Mullin said.
For context on who's offering this assessment: DHS employs more than 200,000 workers and runs an annual budget north of $100 billion. Mullin's own explanation for the fragility went further than the AI theory. "We just have a lot of technologists thrown at us and our world runs on technology now," he said. "We used to have backup systems. And the backup systems are overwhelmed by the new systems now."
He also told Fox viewers to "breathe," and put his confidence that the US is not under a cyber attack at "98%."
The grounding itself was no small event. The FAA halted all air travel in the US for several hours on November 13, with Transportation Secretary Sean Duffy appearing alongside Mullin at a news conference that day.
What does this actually change for operators?
For practitioners running IT infrastructure, the honest answer is: not much yet — and that's the point. No agency has produced evidence tying the outages together. The government's top security official is working from a single informal conversation. If you were waiting for a federal diagnosis of whether AI deployments are destabilizing critical systems, you're not getting one this quarter.
The underlying risk, though, is real and long-predates the AI boom. Technologists have warned for years that America's digital and physical infrastructure is rickety, and that isolated failures cascade. The 2003 Northeast blackout is the canonical case: New York City and big swaths of the country lost power because of a software glitch in Ohio.
Now layer AI on top — a technology whose own builders, including Anthropic's Dario Amodei on interpretability and OpenAI following the Hugging Face incident, have said they don't fully understand how it works or how to control it.
Journalist Josh Tyrangiel, writing in The Atlantic, put the operating environment bluntly: "We should plan to wake up in Software Beirut, where nothing is broken all the time, but anything might break at any time."
The signal-to-noise problem
The confusion cuts both ways. On Tuesday, one reporter was booked to discuss "the November 17 blackout" — an event that never happened, originating from a debunked social media post about a supposed 90-day blackout starting November 17. Hoaxes now travel through the same channels as genuine incident reporting.
That leaves two practical takeaways for anyone managing systems or covering them:
- Treat post-incident explanations from officials with skepticism until they're backed by more than "one guy" — Mullin himself conceded he'd like more information from technology experts in his own organization.
- Accept that sometimes things break because things break, and sometimes a coincidence is a coincidence. Sometimes it's even a squirrel.
For businesses, the working assumption heading into next year is chronic instability rather than a single fixable cause. Backup systems, in Mullin's own words, are already "overwhelmed by the new systems" — which means redundancy plans built for the pre-AI stack may need a fresh look, regardless of whether AI turns out to be the culprit this time.
Source: feeds.businessinsider.com
More from Nathan Brooks
Show full bio
News editor covering regulation, policy and their impact on entrepreneurship, startups and venture funding at Business Bearings.
2,352 articles