The Pentagon Has M-Code. It Can't Use It.

The U.S. military has a satellite problem it can't command. Nine GPS III satellites are in orbit right now, broadcasting a jam-resistant military signal called M-code — and the ground system designed to operate them in full is either getting canceled or stitched into a 2016 stopgap that was never supposed to survive this long. The decision lands March 31, and the irony is so thick you could fuel a rocket with it.

The Pentagon is weighing termination of Raytheon's OCX (Operational Control System) contract after a delay that now stretches past seven years, with costs ballooned from an originally promised $1.5 billion to somewhere between $6 billion and $8 billion, depending on which government accounting document you trust. The company's current contract option expires March 31 and is unlikely to be extended in full, according to SpaceNews. The milestone decision authority, Michael Duffey, the Pentagon's acquisition chief, has the findings from Space Force's analysis on his desk.

OCX was supposed to be the modern ground segment for GPS — the software layer that talks to the satellites and delivers M-code capability to troops. M-code is the hardened, jam-resistant signal that makes GPS useful in a contested electronic warfare environment. Without it, you're flying blind in the ways that matter most. The Government Accountability Office has warned in back-to-back annual reports that these delays increase the risk that U.S. and allied warfighters will be unable to conduct successful operations in contested environments due to lack of access to modernized GPS position, navigation, and timing information. GAO's most recent report flagged it. So did the one before that.

The weird part is that Space Force accepted delivery of a mission-capable OCX system in 2025 and assumed operational control. That sounds like good news. It isn't, really. The delivered system is OCX Block 1/2 — currently in testing — and it has problems. Thomas Ainsworth, the acting Space Force acquisition executive, testified at a March 25 House Armed Services Committee hearing that there were "problems in program management, problems with the contractor performance, problems in system engineering, both on government and on the contractor side, over a number of years." That's a remarkably clean summary of a very messy situation. Rep. Don Bacon, a former Air Force officer, called the program's history what it is.

Air Force Secretary Troy Meink was blunter. "Like, 15 years ago, OCX was having challenges. And it continues to do so," he said at a conference in March. Fifteen years.

The satellites are not the problem. Lockheed Martin has launched nine GPS III satellites as of January 27, 2026, with a tenth expected this month. The constellation — 31 satellites total, comprising eight GPS III, 11 GPS IIF, and 12 older GPS IIR/IIR-M birds — is healthy. GPS III offers roughly eight times better jamming resistance than its predecessors. The spacecraft work. The ground doesn't.

The stopgap is called AEP (Architecture Evolution Plan), and Lockheed Martin contracted it in 2016 specifically to operate newer GPS III satellites while OCX kept slipping. AEP was always a workaround. It can handle the newer satellites in a limited mode, which is better than nothing. But it cannot deliver M-code. It cannot do what OCX was built to do. And now the talk inside Space Force is about whether to graft OCX's software into AEP — essentially abandoning the big-bang modernization approach for a piecemeal integration of code that was supposed to replace the whole thing.

The question nobody's fully answering is what happens to the 22 GPS IIIF satellites on order. OCX 3F — the upgrade needed for the follow-on GPS IIIF constellation — isn't expected to be ready until fiscal 2027 at the earliest, with operational acceptance slipping to 2028, per SpaceNews. And since OCX 3F builds on the OCX software baseline, any further delays to OCX itself correspondingly affect OCX 3F's timeline. The interdependencies don't get easier the longer you wait.

Meanwhile, the DoD is facing a potential gap in the availability of GPS receiver cards — some current cards can no longer be produced and their M-code-capable replacements have been delayed, per GAO. The satellites are up. The ground is lagging. The user equipment is lagging worse. Three moving parts, none of them synchronized.

GAO director Jon Ludwigson, overseeing defense acquisitions, called it what it is: "This has clearly been a beleaguered program. I think folks would have very much liked it to have been completed and called a win." You can hear the exhaustion in that quote. Seventeen years of liking that, by the way.

What's being decided in the next few weeks is essentially whether to keep paying Raytheon to finish OCX as designed, or to cut bait and pour OCX's code into a system it was supposed to replace. Neither option is clean. Terminating means a potential gap in capability and a new procurement cycle for whatever comes next. Staying the course means paying more and hoping the defects get fixed — a hope that has not historically been rewarded.

The underlying issue is that GPS is critical infrastructure nobody thinks about until it breaks. Seven hundred-plus weapon systems depend on it. The people who depend on it most are the ones least able to file a bug report.