ComputerWeekly

US tech outages spark conversation on resilience in the age of AI


A series of widespread technology outages affecting a wide range of organisations across the US has sparked a discussion around business continuity and cyber resilience, and speculation over the impact of artificial intelligence (AI) systems.

From downtime affecting the likes of Apple, Microsoft and Workday, to a nationwide ground stop affecting all flights in the country after a Federal Aviation Authority (FAA) networking hardware upgrade went awry, and from interruption to consumer banking services to a confirmed network outage at the National Security Agency (NSA), the IT problems have hit home across the US.

Although the disruption has not been publicly confirmed as the result of a cyber security incident or incidents, ordinary members of the public are naturally asking the question, “are we under attack?”.

Que sera sera

For now, the question of whether the outages are a merely unfortunate coincidence or the result of coordinated attacks on US infrastructure by a hostile power is the subject of some debate in mainstream discourse.

In response to the outages, president Trump, questioned on the situation ahead of an Asean summit in the Phillipines, was quoted as saying: “It’s life. When you’re big and powerful and have all the money, things like that happen.”

But this said, the wider US government has been crystal clear on its view; interviewed earlier this week, homeland security secretary Markwayne Mullin said that he was virtually certain the outages were not related to a cyber attack.

Appearing before the House Homeland Security Committee on Wednesday 18 November, director of national intelligence Jay Clayton took a similar approach. He said: “We have no indication that these incidents are related, that they’re coordinated. We also have no indication that they’re result of a state-sponsored malign actor.”

In his interview, Mullin speculated that artificial intelligence (AI)  – which he termed super intelligence (SI) in line with directives from the president – could have played a role, and other politicians, such as Virginia senator Mark Warner, also pointed to the impact of AI.

Writing on social media platform X, Warner said that as AI made it possible to identify more issues across more systems more quickly, organisations were coming under pressure to fix them quickly and often all at once, putting strain on systems that were never designed to handle such rapid iteration.

“While we still need to get to the bottom of these incidents, the evidence so far points toward legitimate efforts to secure our systems, not malicious activity,” said the senator.

“Still, these disruptions should serve as a wake-up call. It feels like every day brings another reminder that AI is moving faster than our ability to manage its consequences.”

So are the authorities on the right track? And is AI’s impact a factor in the way that senator Warner supposes?

Steve Povolny, vice president of AI strategy and security research at Exabeam, said it was most likely the incidents were indeed a coincidence.

“[This] sounds like a cop out, but probably isn’t,” he said. “The services on that list share a lot of plumbing underneath: the same cloud providers, DNS, certificates and networking vendors. One bad update or failed piece of equipment can take down names that look unrelated…. People also notice outages in clusters and forget the quiet months.”

Povolny rejected outright the idea of a coordinated attack by nation state actors from China, Iran or Russia. “[It] would be the loudest thing a state actor has ever done, and attackers who want access usually work to keep it quiet,” he said.

Turning to the impact of AI, Povolny said there was a possibility that someone using an AI tool to move faster, or an automated change pushed live without review, could have been a factor in some of the disruption.

Security lessons in the making

While not a cyber security incident in the traditional sense of the term, there are still security lessons to be drawn from the current wave of outages, said Halcyon Ransomware Research Center vice president Cynthia Kaiser, a former FBI operative.

“We are in an era where AI is accelerating the discovery of legacy software flaws, forcing organisations to race to patch vulnerabilities before adversaries can exploit them, which inevitably creates friction and disruptions,” she said.

“Even when an outage isn’t caused by a cyber attack, these moments of operational distraction leave dangerous gaps open for threat actors and opportunists to take advantage of, making it crucial for both institutions and everyday individuals to remain on high alert against social engineering and unsolicited outreach.”

ISC2 chief operating officer Casey Marks echoed the sentiment, saying that the so-called CIA Triad – confidentiality, integrity and availability – a foundational model for the development of security systems, already recognised that keeping critical systems accessible was as essential to effective cyber security as data protection, something that the outages have clearly highlighted.

Marks recommended that organisations take a good look at their business continuity and recovery plans.

“Start by identifying the systems your organization can’t operate without and the vendors those systems depend on. Then test what happens when those services go down. Teams should know who can make recovery decisions, how to reach employees and customers if their usual channels are down, which operations can continue, and which systems need to come back first,” said Marks.

“A technical failure can stop a business just as effectively as an attack, even when no data is stolen. Planning for disruption and testing how you’ll recover deserve the same attention as preventing attacks,” he said.

Povolny also highlighted the similarities between a technical outage and a cyber intrusion, saying that in the throes of a crisis it can be hard to tell the two things apart.

“Services go dark, people can’t log in, and nobody yet knows whether the cause was a mistake or a break-in. The teams that sort it out fastest are the ones who already know what normal looks like for their systems and accounts. When a platform fails, they can check right away whether a privileged account did something unusual just before it,” he said.

“The teams that can’t answer that tend to either call an attack too early or wave it off too late, and both cost them. Resilience matters here too. If a company has never practiced running without its main network or cloud provider for a day, the first real outage is a poor time to learn how.”

But Vasu Jakkal, chief strategy officer at Armadin – herself a former Microsoft security leader – concluded on a more upbeat note.

“While the disruption to our day-to-day lives is a concern and there is work to be done, I’m optimistic because it means that defenders are taking the threat of AI-driven attacks extremely seriously and IT teams are working on modernising infrastructure for an agentic world,” she said.

“These organisations have recognized they need to upgrade their systems because attackers aren’t waiting until they test updates to strike. But these past few days have demonstrated that defenders are in an impossible position: either they accept the risk that hackers will exploit vulnerabilities before they’re ready to update their systems, or they rush patches faster. If given the option, they should choose the latter every single time.”

Describing the outages as growing pains, she said there was unfortunately a high likelihood of more such disruption to come.

“I’m confident that we’ll move past these hiccups. It’s a process, but a necessary one to protect ourselves long-term,” said Jakkal.



Source link