Grounded by a Legacy System

United Airplane on the Tarmac with passengers exiting

Grounded by a Legacy System

United Airlines’ August 6, 2025 Computer Outage

By Mary Keane, CyberWyoming Cybersecurity Business Counselor with assistance from AI

On August 6, 2025, I was about to board a flight from Portland, Oregon, to Denver. The plane was scheduled to leave at 6:16 p.m. Eastern, but boarding came to a screeching halt. United Airlines’ computer system had gone down nationwide.

From Newark to Denver to Portland, planes were grounded and passengers stranded. What happened wasn’t a cyberattack, as many first feared. Instead, the disruption came from deep inside United’s own infrastructure: a failure of its Unimatic system.

What Is Unimatic?

Unimatic is a legacy mainframe, first deployed in 1997, that manages critical flight data:

  • Weight and balance calculations (required before every takeoff)
  • Flight timing and scheduling data
  • Information feeds to dependent operational systems

In short, when Unimatic goes down, United Airlines can’t safely fly.

How the Outage Unfolded

Shortly after 6 p.m. Eastern, Unimatic failed, causing dependent systems to stop functioning. Without accurate weight and balance data, United had no choice but to halt departures at major hubs, including Newark, Denver, Houston, San Francisco, and Chicago.

By the end of the evening:

  • Over 1,000 flights were delayed (about 34–35% of United’s schedule).
  • Roughly 200 flights were canceled (about 6–7% of the schedule).
  • Recovery began around 10 p.m. ET, though ripple effects lasted into the next day.

United later described the incident as a “technology disruption” and emphasized it was not cyber-related. The FAA and U.S. Transportation Secretary confirmed the outage was isolated to United’s systems.

Legacy Systems Under Strain

United hasn’t disclosed the precise trigger—whether a software bug, hardware failure, or data corruption. What is clear: Unimatic is already slated for replacement. Nearly three decades old, it’s a relic of the 1990s still carrying the load of a modern global airline.

The outage underscored the risk of relying on outdated technology in industries where safety and timing are non-negotiable. As The Wall Street Journal noted, modernization is underway, but the urgency is undeniable.

My Experience

At first, the flight crew thought the problem would be resolved quickly. They unloaded the passengers who had boarded, and we waited at the gate for another two hours. Eventually, we reboarded, only to sit on the tarmac for 90 more minutes while Denver tried to juggle incoming flights. Aircraft already in the air during the outage were all trying to land at once.

I checked my watch, wondering if our flight crew would run out of duty time. Five minutes later, the captain confirmed it: they had “timed out.” Our flight was canceled.

By the time I got off the plane, there were 200 people in line for two gate agents. Instead of waiting, I opened the United app: within minutes I rebooked for the next day, selected a hotel with a shuttle, and downloaded a QR code for a $15 meal voucher. I showed the women in front of me how to do the same.

Luckily, I hadn’t checked luggage. Everyone else had to wait—baggage handling had also been affected by the outage.

The next day I flew home in a middle seat (all that was left when I rebooked). I considered myself fortunate. It was the end of my trip; I only had to cancel a few client meetings. But I thought about other passengers—those who may have missed weddings, family milestones, or critical business commitments.

Closing Thought

In the end, United treated the outage as a “controllable delay,” covering some costs for passengers. But the broader lesson is clear: critical infrastructure must be modernized before failures cascade.

For passengers and airlines alike, the August 6 outage was more than an inconvenience. It was a wake-up call. When technology works, like using the United app to rebook my flight, it’s a wonderful convenience. However, the United system outage shows how our world is so digitized that it is easy to take for granted that computer systems ‘just work.’ That’s when we are reminded that when computers break down, there’s often a physical consequence. In this case, I, thousands of United employees, and United staff paid the price.

Listen to This Blog as an AI Podcast

This podcast is an AI-generated audio version of the blog above. It was created using NoteBookLM, an AI tool that transforms written content into natural-sounding spoken narration. The text of the blog is processed by the AI, which then generates a podcast-style recording so you can listen instead of read.

Share:

Register to Receive the Tech Joke of the Week!

This Week's Joke:

How many programmers does it take to change a light bulb?

None, it is a hardware problem!

More Posts: