A nationwide US air travel disruption last week and recent system outages at Microsoft, JPMorgan Chase and United Airlines have prompted questions about whether newly deployed AI software played a role. The US homeland security secretary raised that possibility in a television interview but offered no supporting evidence. Until the causes are established, companies and public agencies cannot determine whether the problems stemmed from AI deployments, aging infrastructure or unrelated incidents.
Officials have not produced evidence linking outages to AI
The secretary said the suggestion came from someone he considered knowledgeable about technology and that it sounded plausible. He also acknowledged he was not certain. No public information has shown that last week’s nationwide aviation disruption and the recent outages at Microsoft, JPMorgan Chase and United Airlines had a common cause, or that new AI software contributed to them.
He said society’s growing reliance on technology may leave older backup systems struggling to keep pace with the complexity of newer systems. He put his confidence that the US was not under cyberattack at about 98%, but that assessment does not establish what caused the aviation disruption or the other outages. For airlines, financial institutions and cloud providers, identifying the root cause is essential to restoring services, maintaining business continuity and assessing risk. Speculation is not a substitute for a technical investigation.
Legacy infrastructure and interconnected systems complicate investigations
Concerns about digital systems’ resilience predate the current AI boom. Technology experts have long warned that parts of the US’s digital and physical infrastructure are poorly maintained, and that a seemingly isolated fault can spread through interconnected systems. The 2003 Northeast blackout is a frequently cited example: a software failure in Ohio contributed to a cascade that cut power across a wide area, including New York City.
AI adds another layer of complexity. Developers are still working to understand how models behave and how best to control them. Connecting new software to existing IT environments can make interactions between systems harder to predict. But complexity alone does not show that AI caused any particular service disruption. Whether the aviation shutdown and corporate outages were connected remains a question for investigators.
Separate confirmed facts from speculation
The public and the media are right to ask what caused major outages, but vague explanations offered after an event should not be treated as conclusions. At the same time, quickly attributing each service disruption to a cyberattack or AI could keep the public on constant alert and make genuine warnings easier to ignore. Waiting until every risk is fully understood before issuing a warning, however, may leave little time to prepare. Clear communication should distinguish confirmed facts from open questions and speculation.
Similar confusion has surfaced on social media. Some users described November 17 as the date of a nationwide blackout, but the claim has no factual basis; there is no evidence that a so-called “November 17 blackout” occurred. Repeating an unverified claim, even to debunk it, can expose more people to it and lead some to believe it.
For now, there is no specific action ordinary users can take in response to an unconfirmed link between these incidents. The practical questions are whether individual services have been restored, how many people were affected, and when operators and investigators will disclose their findings. Whether AI was involved in the aviation disruption or the outages at Microsoft, JPMorgan Chase and United Airlines remains unproven.