Incident sheet · /infrastructure
- System
- Infrastructure & Cloud
- Filed
- Read
- 4 min · 735 words
- Severity
- Routine
NSA's Classified Intranet Went Dark Tuesday, Triggering 'Code Yellow'
NSA Net, the agency's classified intranet linking Fort Meade to cryptologic hubs worldwide, went down Tuesday. A 'code yellow' followed — no cyberattack suspected, but the week's outage streak is brutal.

Event log
- NSA Net, the agency's classified intranet, suffered a serious outage on Tuesday, triggering a 'code yellow' — the second-highest of three alert tiers.
- The outage followed a week of failures: Microsoft 365 down nearly 24 hours Thursday, an FAA ground stop Friday canceling 17,000+ flights, and a JPMorgan software glitch Monday.
- Officials, including DHS Secretary Markwayne Mullin, said there is no evidence of a cyberattack.
- The disruption did not affect secure intelligence communication with President Trump, traveling in Asia.
- Critics link the failures to federal civil service cutbacks across national security and cyberdefense agencies.
The National Security Agency's classified internal network — NSA Net — went down Tuesday, prompting the agency to declare a "code yellow," the second-highest alert in its three-tier system. CNN reported the outage first, and multiple sources familiar with the matter confirmed it.
By Tuesday afternoon, investigators had found no signs of a cyberattack. But the incident landed in the middle of the worst stretch of U.S. infrastructure failures in recent memory, and it raised an uncomfortable question about the systems the government depends on.
What exactly went down?
NSA Net is the agency's official classified intranet. It carries classified internal communications from NSA headquarters at Fort Meade, Maryland, to the country's four primary cryptologic hubs: NSA Colorado in Denver, NSA Georgia at Fort Gordon, NSA Hawaii on Oahu, and NSA Texas in San Antonio.
The network also connects to specialized sites, including the massive Utah Data Center in Bluffdale and NSA Alaska at Joint Base Elmendorf-Richardson. It supports Cryptological Services Groups and agency personnel stationed at major combatant commands and U.S. military installations worldwide.
The system runs on dedicated fiber-optic and satellite channels, deliberately isolated from the public internet. When localized disruptions hit, backup centers around the country are designed to absorb the traffic and keep operations running. On Tuesday, that architecture wasn't enough to prevent a serious outage.
Was the president cut off?
No. Officials stated the disruption will not impact the agency's ability to communicate secure intelligence to President Donald Trump, who is traveling in Asia. He's meeting Chinese President Xi Jinping and Russian President Vladimir Putin at the first-ever trilateral summit in Shenzhen, China.
The NSA also instituted staggered arrival times for employees while it investigated. There's a Five Eyes dimension, too: NSA Net is cleared to support signals intelligence collaboration with the UK's GCHQ, the Australian Signals Directorate, Canada's Communications Security Establishment, and New Zealand's Government Communications Security Bureau, under releasability rules including REL TO USA AUS CAN GBR NZL.
What does this change for everyone else?
For enterprise and government IT leaders, the week's lesson is about concentration risk. Consider the timeline:
- Last Thursday: A Microsoft 365 outage knocked out Outlook, Teams, and Copilot for nearly 24 hours. Microsoft blamed a software defect combined with high traffic and reduced server capacity. Government systems, major defense contractors, and key U.S. allies took the hit — a measure of how heavily U.S. defense and diplomacy lean on Microsoft infrastructure.
- Same day: Workday's cloud platform, used by medium and large organizations for HR, payroll, financials, and planning, suffered a disruption that cut off access to HR and payroll tools.
- Friday the 13th: The FAA ordered a rare nationwide ground stop, halting all U.S. departures for about four hours — the longest such stop since September 11, 2001. Overloaded routers during contractor equipment maintenance caused it. An estimated 300,000 to 400,000 passengers were stranded at the height of the chaos Friday, and more than one million were affected across the weekend.
- The flight math: More than 9,000 flights were canceled Friday, over 6,400 on Saturday, and more than 900 on Sunday — pushing total cancellations past 17,000. Airlines were still repositioning displaced crews and aircraft on Tuesday.
- Monday: United Airlines logged a separate systemwide tech issue delaying departures. JPMorgan Chase, the largest U.S. bank, suffered a major disruption caused by a software bug in on-premises hardware, with ripple effects hitting Bank of America, U.S. Bank, Wells Fargo, and Zelle peer-to-peer transfers.
That's a cloud productivity suite, an HCM platform, national airspace, a global airline, and the biggest bank in the country — all inside six days.
Is anyone pointing fingers?
Department of Homeland Security Secretary Markwayne Mullin said there's no evidence the NSA Net outage or the concurrent nationwide disruptions resulted from a cyberattack. Critics disagree on the cause, not the impact: they suggest the failures could stem from federal cutbacks.
The glitches have coincided with a wave of civil service departures and budget cuts across national security and cyberdefense agencies, plus recent high-level leadership changes at the NSA and U.S. Cyber Command.
No smoking gun yet. But when a network built on dedicated fiber and satellites, physically separated from the public internet, still goes dark — the redundancy plans deserve a second look.
via news.clearancejobs.com (Source)
More from Olivia Hart
Show full bio
News editor covering money, funding and market shifts around enterprise information technology and the leaders who run at CIO Monitor.
207 articles
The newsroom digest
One email a day with the stories that matter. No noise.