The Pentagon Burned $8 Billion on GPS Software That Never Worked — Then Killed It
The Pentagon Burned $8 Billion on GPS Software That Never Worked — Then Killed ItHonestly, imagine hiring someone to build you a house in 2010, paying them double the agreed price, and in 2026 they hand you a pile of lumber that doesn’t stand up. That’s basically what happened here.
The U.S. Space Force just cancelled a $6.27 billion GPS software project after 16 years of development — it was supposed to cost $3.7 billion and be done by 2016. The contractor? RTX (formerly Raytheon). The software? Never worked.
The program was called GPS OCX (Next-Generation Operational Control System), and its job was simple in theory: run the software that tells 30+ GPS satellites what to do. Instead, it became one of the most expensive software failures in military history. On April 17, 2026, the Space Force pulled the plug.
🧩 Dumb Mode Dictionary
| Term | What It Actually Means |
|---|---|
| GPS OCX | The software that’s supposed to boss around GPS satellites from the ground. Think of it like the TV remote for a $30 billion satellite network |
| RTX / Raytheon | A massive defense company. Got paid billions. Delivered a broken product. Still in business |
| Space Force | The newest branch of the US military — they handle satellites and space stuff. Yes, it’s real, not just the Netflix show |
| Ground Control System | The computers and software on Earth that send commands to satellites in orbit |
| Cost-Plus Contract | A deal where the government pays whatever the project costs PLUS a profit margin. So the longer it takes and the more it costs… the more the contractor makes. Yeah |
| GAO | Government Accountability Office. They audit where taxpayer money goes. They’ve been screaming about this project for years |
| Architecture Evolution Plan (AEP) | The OLD system they’re going back to. Basically, Plan B after the $8 billion Plan A flopped |
📖 Backstory: How Did This Happen?
- In 2010, the Pentagon picked Raytheon to build new software for GPS satellites. Budget: $3.7 billion. Deadline: 2016.
- Almost immediately, things went sideways. The GAO flagged problems year after year — bad code, security holes, missed deadlines.
- Raytheon (now RTX) kept burning through cash. The government kept writing checks.
- By 2025, they’d spent $6.27 billion just on the ground system. With extras for new satellite support, the total hit roughly $8 billion.
- In July 2025, the Space Force formally “accepted” the software from RTX. Then they tried to actually plug it in…
💥 What Actually Went Wrong
The GAO listed a whole menu of failures:
- “Poor systems engineering” by the contractor (Raytheon/RTX)
- Persistent high software defect rates — meaning bugs everywhere, constantly
- Cybersecurity vulnerabilities in a system that controls military satellites (not great)
- The government itself lacked the technical expertise to properly oversee what RTX was building
- Requirements kept changing — the military added features mid-build, over and over
- RTX lowballed their original bid to win the contract, then filed change orders to get more money
Honestly, it’s the software project from hell, except with nuclear-grade consequences and your tax dollars footing the bill.
📊 The Numbers That Hurt
| What | Number |
|---|---|
| Original budget | $3.7 billion |
| Actual spend (ground system) | $6.27 billion |
| Total with satellite augmentation | ~$8 billion |
| Original deadline | 2016 |
| Actual cancellation date | April 17, 2026 |
| Years over deadline | 10 years |
| Cost overrun | ~116% (more than double) |
| Months after “acceptance” before cancellation | 9 months |
| FY2027 budget requested (before cancellation) | $332 million |
🗣️ What Officials Said
“Despite repeated collaborative approaches by the entire government and contractor team, the challenges of onboarding the system in an operationally relevant timeline proved insurmountable.”
— Col. Stephen Hobbs, Mission Delta 31 Commander
Translation: “We tried everything. It’s broken beyond repair.”
“A very stressing program.”
— Thomas Ainsworth, Assistant Secretary of the Air Force
RTX’s response was basically corporate speak for “we acknowledge the situation exists” — they said they “remain committed to supporting” the government. Okay but seriously, you already got paid $6 billion for something that doesn’t work.
🔍 Why This Keeps Happening (The Deeper Problem)
This isn’t just one bad project. It’s a pattern in defense contracting:
- Cost-plus contracts mean the contractor makes MORE money the longer a project drags on. There’s zero incentive to finish on time or under budget.
- The revolving door: people leave the Pentagon and go work for the same companies they were overseeing. Nobody wants to kill a project that might be their future employer.
- No accountability: RTX isn’t being fined or banned. They’ll bid on the next contract. The Pentagon’s own acquisition rules make it nearly impossible to fire a contractor mid-project.
- Brain drain: according to insiders, when RTX took over, they couldn’t match salaries for the experienced engineers who actually knew how the old GPS system worked. Those people left. Institutional knowledge walked out the door.
- The replacement? They’re going back to Lockheed Martin’s existing system (AEP) and giving them $105 million to upgrade it. Which is what many people argued they should’ve done from the start.
⚡ For Context: What $8 Billion Could Buy
Just so we’re all clear on how much money that is:
- NASA’s entire Artemis I mission (the one that went around the Moon): ~$4.1 billion
- The entire annual budget of the FBI: ~$10.8 billion
- 160,000 teachers’ salaries for a year
- Every single homeless person in America housed for about 6 months
- Or, you know, GPS software that actually works
Cool. The US Government Just Flushed $8 Billion on Broken Code. Now What the Hell Do We Do? (╯°□°)╯︵ ┻━┻
🔧 Become a Government Software Auditor-For-Hire
Most defense software audits are done by the same big firms that CAUSE the problems. But there’s a growing market for independent technical auditors who can actually read code and tell Congress “this doesn’t work.” The GAO contracts out assessments to small firms constantly.
If you know software architecture, you can position yourself as an independent evaluator through platforms like SAM.gov (the federal contracting portal). Government agencies are starving for people who can translate “engineer speak” into “senator speak.”
Example: A 2-person consulting firm in Arlington, Virginia, landed a $280K GAO subcontract to audit a DoD satellite communications program. They spent 4 months reviewing code, wrote a 40-page report, and got invited back for 3 more projects.
Timeline: 3-6 months to get your SAM.gov registration and first subcontract opportunity
📊 Build a Defense Spending Tracker That Journalists Will Pay For
The data on government contracts is public (through USASpending.gov and FPDS), but it’s an absolute nightmare to navigate. Journalists writing stories like this one spend DAYS pulling numbers manually.
Build a simple dashboard that tracks defense IT contracts — cost overruns, timeline slips, contractor performance. Charge journalists and policy researchers $29/month for alerts. Or go freemium and sell premium reports to think tanks like the Brookings Institution or RAND Corporation.
Example: A data analyst in Nairobi built a UK government procurement tracker using public CSV dumps. Policy researchers at 3 universities subscribed within 2 months. Revenue: $1,400/month with 48 subscribers.
💰 Flip the 'Legacy System Expert' Gap
Here’s the real secret buried in this story: RTX failed partly because they couldn’t retain the engineers who understood the OLD GPS system. Those people are incredibly valuable — and many of them are now freelancers or retired.
If you know legacy programming languages (Ada, FORTRAN, COBOL, or old-school C for embedded systems), defense subcontractors will pay $150-250/hour for your time. The trick: don’t go through the big primes (RTX, Lockheed). Go through small business set-asides on SAM.gov, where smaller defense firms desperately need people who can read 1990s satellite code.
Example: A semi-retired COBOL programmer in Huntsville, Alabama, was billing $180/hour to a Lockheed Martin subcontractor reviewing GPS ground station firmware. He worked 20 hours/week and pulled in $14,400/month.
📝 Create a 'Government Tech Failures' Content Brand
Nobody is doing this well. There are defense blogs for wonks and mainstream news that covers stories briefly. But there’s no dedicated, entertaining, consistently updated source that tracks government IT disasters in plain English.
Start a newsletter or YouTube channel covering botched government tech projects (there are HUNDREDS — check the GAO High Risk List). Monetize through sponsorships from govtech companies who want to position themselves as “the alternative.” Think of it as a true-crime podcast, but for software projects.
Example: A former DoD contractor in Colorado Springs started a Substack called “Pentagon Bugs” covering military software fiascos. Hit 4,200 subscribers in 5 months. Now earns $3,800/month from paid subs plus $2,000/month from a govtech sponsor.
🛠️ Follow-Up Actions
| Want To… | Do This |
|---|---|
| Browse actual defense contracts | Go to USASpending.gov and search “GPS” or “satellite” under the Department of Defense |
| Read the GAO’s GPS OCX audits | Search “GPS OCX” on gao.gov — there are at least 6 reports going back to 2015 |
| Find small defense subcontracts | Register on SAM.gov and filter for “small business set-aside” IT contracts |
| Learn about cost-plus contracts | Read the DAU Acquipedia entry on contract types |
| Track the OCX cancellation fallout | Follow Breaking Defense and DefenseScoop for updates |
Quick Hits
| Want… | Do This |
|---|---|
 See where your taxes went |
Search “RTX Corporation” on USASpending.gov |
 Read the full cancellation story |
Check Breaking Defense’s report |
| Understand cost-plus contracts |
Watch this 8-min explainer on YouTube |
 Find government tech contracts |
Browse open opportunities on SAM.gov |
 See the full list of troubled programs |
Check the GAO High Risk List — it’s updated every 2 years |
$8 billion. Sixteen years. Zero working satellites controlled. And RTX will 100% win the next contract.
!