Story
Nov 20, 2026

The Outages Weren’t Proof of an AI Attack—but They Exposed a Bigger Risk

The outages have sparked competing explanations: some see a dangerous rush to patch flaws AI can find, while others warn against treating coincidence as evidence of an AI-driven attack. Across the debate, a shared concern persists—critical systems may be too fragile to absorb rapid change.

Over five days, disruptions hit Microsoft services, U.S. air travel and major banks. A sharply different interpretation came from the author of one security-industry analysis: overwhelmed systems and undisclosed root causes, they argued, warranted investigating the possibility of denial-of-service attacks—not simply accepting early statements that there was no attack.“If it looks like a DDoS attack”

Homeland Security Secretary Markwayne Mullin said an unnamed industry contact suspected AI systems writing or fixing code could be overwhelming older systems. But Mullin also said he was 98% certain the outages were not a cyberattack and described them as individually unrelated.“The Trump White House is starting to admit AI is breaking the internet” Other experts urged caution: one called the failures separate incidents and said cascading outages were “nothing to do with writing code” using AI.“Why the internet suddenly broke all at once”

As the week unfolded, officials and executives increasingly pointed to a different AI connection: tools may be uncovering vulnerabilities faster than companies can safely fix them. Cybersecurity CEO Kevin Mandia said updates pushed at a pace that leaves little time to test outcomes can break systems; Senator Mark Warner warned that rushed patches could mean more outages.“AI was probably involved in recent tech outages” A separate GAO report raised another concern about AI: testers bypassed chatbots’ safeguards to obtain plans for attacks, though the agency cautioned that its limited tests were not a statistically significant sample.“Washington convinces 4 AI chatbots to plan terror attacks…”

Cybersecurity journalist Nicole Perlroth called the instability an “opening chapter,” while noting that vulnerability counts do not prove AI caused their rise. Others stressed that aging infrastructure and complex systems can fail during routine changes, even without an attacker.“‘This, I’m sorry to say, is just the opening chapter’”“These tech outages are giving Y2K, AI agent style”

The political argument widened, too: former CISA director Jen Easterly said federal staffing and budget cuts had eroded the agency’s capacity and trust with partners.“Widespread Tech Outages Put Focus on Trump’s National Security Cuts” The practical advice from another analysis was less dramatic than the competing theories: prepare for disruptions, strengthen systems and avoid panic.“Trust, but prepare: How to make sense of AI's age of upheaval”

Story coverage