Outages expose vulnerabilities in US digital infrastructure


A series of major computer outages brought critical infrastructure in the United States to a halt over roughly a week, affecting air traffic control, financial transactions, national intelligence networks, corporate work platforms and consumer digital services. The disruptions to critical U.S. systems began with a Microsoft (MS) cloud outage on the 12th and continued over six days, affecting the Federal Aviation Administration (FAA) on the 13th, major banks on the 16th, and the National Security Agency (NSA) and Apple services on the 17th, local time.

People wait in Terminal B at Orlando International Airport in Florida, U.S., on the 13th due to a system outage. /Courtesy of Yonhap News Agency
People wait in Terminal B at Orlando International Airport in Florida, U.S., on the 13th due to a system outage. /Courtesy of Yonhap News Agency

On the 17th, the National Security Agency, a key intelligence agency that protects government communications networks, suffered an outage of its internal classified communications network, prompting employees at its Fort Meade, Maryland, headquarters to delay reporting to work. The same day, an update to Apple’s internal systems was immediately followed by outages across consumer services including iCloud login, the App Store and Apple Pay, as well as its virtual private network (VPN) for remote workers and its internal portal.

Earlier, on the 12th, Microsoft’s workplace platforms Outlook and Teams went down for two days. The following day, a Federal Aviation Administration communications network failure halted all aircraft departures across the United States for 4 hours and 30 minutes, the longest such shutdown since the Sept. 11, 2001, terrorist attacks. On the 16th, online transactions at major banks including JPMorgan Chase, Bank of America and Wells Fargo stopped working.

The five organizations affected by the outages are vastly different, ranging from public security agencies to private tech giants with some of the world’s most advanced technology. The causes also varied. At the Federal Aviation Administration, aging communications equipment was the source of the problem. At Microsoft and Apple, by contrast, errors occurred during the companies’ own work on their latest cloud and internal software systems. Microsoft said, “A bug that had been dormant since July caused a system failure during routine network capacity upgrades and traffic rerouting.” JPMorgan Chase also blamed a software bug in its own servers.

The National Security Agency’s trouble began as it deployed a security patch to block a server remote-control vulnerability, a PHP bug, discovered by artificial intelligence (AI). A National Security Agency Spokesperson said in a statement on the 17th, “We experienced a temporary degradation in performance during scheduled routine maintenance. Some internal systems were affected, but our ability to carry out core missions remains fully operational, and our ability to defend national security is also maintained.”

Experts said the relentless series of computer outages resulted from a mix of old equipment and new software. The United States has the world’s most advanced digital industry, but the public and private operating systems on which that industry relies have not all been built with the latest technology. Instead, because the United States has led the global digital industry for so long, old equipment and the latest software coexist uneasily. Banking, transportation and government computer networks built up over decades since the mid-20th century operate like geological strata, with layers of new software piled on top of old hardware and early programming languages.

For example, COBOL, a programming language created as a stopgap in 1959, still processes more than $1 trillion, or about 1,330 trillion won, in financial transactions a day in the banking sector after decades of minor modifications. Christiaan Beek, a researcher at security company Rapid7, told Bloomberg, “Complex systems fail far more often for boring, routine reasons than because of external adversaries. Worryingly, that means the systems America relies on are fragile enough for us to bring them down ourselves, without an attacker.”

With old equipment and the latest software coexisting uneasily and the systems of different agencies and companies closely connected, changes at one point can upset a delicate balance. Attempts to fix one part also frequently spread outages across a far wider area than expected. The Federal Aviation Administration outage, which caused about 20,000 flight cancellations, resulted from an error during equipment replacement in a rural area that was not contained and overloaded the entire nationwide air traffic control communications network. In the financial sector, a software bug within JPMorgan Chase spread through the payments network, causing a chain reaction that also prevented Bank of America and Wells Fargo customers from using the Zelle money transfer service and their online accounts.

Cynthia Kaiser, a former Federal Bureau of Investigation (FBI) official and vice president at security company Halcyon, told Reuters, “The recent large-scale outages across industries show just how deeply interconnected America’s digital infrastructure is. They show how the same technical error or failed software update can have enormous effects across critical systems.”

Wiring for Amazon Web Services (AWS) Trainium3 UltraServers operating at a quality assurance (QA) lab in Austin, Texas, U.S. /Courtesy of Yonhap News Agency
Wiring for Amazon Web Services (AWS) Trainium3 UltraServers operating at a quality assurance (QA) lab in Austin, Texas, U.S. /Courtesy of Yonhap News Agency

The rapid advancement of artificial intelligence (AI) technology is also accelerating system maintenance at an abnormal pace, fueling clashes between old and new systems. Major U.S. companies and agencies are scouring existing systems for vulnerabilities and rapidly applying security patches at an unprecedented pace to prepare for attacks by hackers using AI as a weapon. U.S. Director of National Intelligence Jay Clayton told Fox Business on the 17th, “Everyone recognizes that it’s time to upgrade systems. But when you carry out an upgrade, the system doesn’t always come back up as expected.”

Some have also pointed to a combination of insufficient investment in backup systems to isolate outages and cuts to skilled staff driven by AI advances as a reason minor maintenance errors can lead directly to systemwide failures, as in the recent incidents. The Federal Aviation Administration, United Airlines and Microsoft are among those that have recently made major staff cuts. Joe Slowik, director of threat research at risk intelligence company Dataminr, told Bloomberg, “We’ve relied too heavily on unstable infrastructure to run more smoothly than it actually does. We’ve neglected investment in redundant backup systems that could intervene to prevent the worst-case scenario, whether it’s an error, a bad update or even a cyber incident.”

※ This article has been translated by AI. Share your feedback here.

Leave a Reply

Your email address will not be published. Required fields are marked *