[
Many people are understandably shaken after a string of nationwide outages over the past week—from flights being grounded, to emails not coming through, bank transactions failing, and much more—with no clear explanation from politicians or businesses about what exactly happened.
One thing officials have made clear is that the outages are not a result of a coordinated cyberattack, quelling fears that AI may have empowered bad actors to take down critical systems. However, AI still appears to have played a significant, if unexpected, role. Advanced AI models are speeding up the discovery of software vulnerabilities, which tech teams are racing to patch before bad actors can exploit them with equally powerful AI tools.
The botched implementation of those patches was linked to an outage earlier this week of an internal communications system at the National Security Agency (NSA), as well as the “banking outages,” a spokesperson for Senator Mark Warner’s team told Fortune. Bank of America, Wells Fargo, and JPMorgan Chase all suffered outages this week.
“It feels like every day brings another reminder that AI is moving faster than our ability to manage its consequences,” Warner said in a statement. The Washington Sun also reported that bank CEOs and the NSA told Congress that AI-driven patches were to blame for the service disruptions.
The irony is that the tools aimed at preventing system outages seem to be causing them, or at least influencing new forms of human error. Bank of America disclosed last month it had hired 1,000 contractors to patch flaws exposed by Anthropic’s Mythos model, on top of 3,000 internal staffers, CNBC reported. That does not mean businesses should abandon AI-powered defensive cybersecurity tools, but rather suggests that their implementation is not going according to plan.
“This is a failure of implementation,” Amit Shah, founder and CEO of InstaLILY AI, tells Fortune. “The AI finds a vulnerability in seconds and may not know what else breaks when you fix it. The person approving the patch sees a recommendation, not the systems it touches. So a tool built to prevent outages can end up causing them.”
Anxiety around “the theoretical risk of a breach” is prompting a “rush to eliminate” vulnerabilities, Jeff Collins, CEO of the IT technology platform WanAware, tells Fortune. “AI now discovers vulnerabilities at machine speed while our testing and deployment safeguards remain stuck at human speed.”
Thanks to AI tools’ thoroughness, IT teams are overwhelmed with fixes to implement, Palo Alto Networks CEO Nikesh Arora said on X. “Some of the recent outages are a consequence of a deluge of patches,” Arora said. Between these teams’ mandates to plug holes, and the threat of “AI-assisted hacks, we have a challenge,” Arora said, with “no silver bullets.”
“The issue is that the volume of issues we’re fixing is so large that it’s pressuring our staff to apply large amounts of fixes that can break things,” Cliff Steinhauer, Director of Information Security & Engagement, National Cybersecurity Alliance tells Fortune. “We don’t always have time to fully test every update because there are so many and attackers can pick up the known vulnerabilities and exploit them.”
But Steinhauer doesn’t consider this an implementation failure. Instead, he sees it as a necessary phase enterprises must go through as they secure their code bases. “I think we are in a bell curve of vulnerabilities discovery and patching. The hope is that we’ll get to a peak, then it will start to descend and slow down once we start to fix the underlying issues in our software.” He compares it to road maintenance. If the Department of Transportation (DOT) “finds a crack in the bridge, they have to shut down traffic to fix it, which is disruptive. But if the bridge collapses, that’s obviously worse.”
Human or AI?
The onus of responsibility between the human and the AI remains hard to pin down. While the human may not have properly vetted the impact of the fixes, it’s unclear what information they were operating on from the AI tool. Did the AI create anxiety by overstating the vulnerabilities’ significance? Did it recommend changes en masse, that perhaps a human approved without proper vetting? Are security team managers pushing teams to embrace AI as fast as possible, to avoid a potentially catastrophic breach?
“The failure isn’t that AI patches too aggressively or that exhausted engineers rubber-stamp fixes, but that we plugged a hyper fast diagnostic engine into legacy operational pipelines never built to absorb that velocity of change,” Collins said.
OpenAI offers cybersecurity-specific models to help organizations find vulnerabilities at scale, and is rapidly enrolling new customers to its Daybreak program that provides approved security professionals with its most powerful defensive cyber tools. The goal is to “patch the planet,” as OpenAI has named its Daybreak initiative to secure code better than ever before. Its Daybreak products include dashboards that classify bugs by severity and aim to help teams triage them. Anthropic’s version of Daybreak is called Project Glasswing.
A week and a half before its outage, on Nov. 5, OpenAI president and co-founder Greg Brockman announced a partnership with JPMorgan Chase to use OpenAI models to help secure its systems.
An OpenAI spokesperson said that the JPMorgan outage stemmed from legacy systems getting overwhelmed by code modernization work. A JPMorgan Chase spokesperson said OpenAI’s tools were not a factor in the outage. The spokesperson said the problem traced back to “a temporary issue related to a software bug in on-premise hardware systems.” The spokesperson added hat the company did not identify that bug through AI tools or fix it with AI.
OpenAI co-founder and president Greg Brockman recently told The Information there was “no question” that cyber would eventually become a bigger business for the company than coding. The company swapped out its Chief Revenue Officer earlier this year for Dali Rajic, and tasked him with leveraging his cybersecurity enterprise sales experience to accelerate that business line.
The widespread adoption of AI tools for defensive cybersecurity might be trickier than these organizations initially expected. Plus, lawmakers are seizing on the recent outages to demand more patch-related guardrails. Representative Andy Ogle plans to introduce legislation today that calls for better coordination as AI speeds up the discovery and exploitation of software flaws, NextGov reports.
https://fortune.com/img-assets/wp-content/uploads/2026/11/GettyImages-2271848081-e1795031217617.jpg?resize=1200,600
https://fortune.com/2026/11/18/ai-played-a-major-role-in-the-nsa-banking-outages-but-not-in-the-way-you-may-think/
Emily Forlini




