GAO told SDA to prove the technology first. It spent billions anyway
Summary
The Space Development Agency's Proliferated Warfighter Space Architecture is a planned constellation of 300 to 500 satellites in low Earth orbit, meant to detect and track missile threats and cost nearly $35 billion through fiscal year 2029. Two GAO reports, about 11 months apart, trace what happened after GAO first flagged the program's core technology as unproven. In February 2025, GAO found SDA's demonstration tranche hadn't fully proven its laser-communications technology -- yet SDA had already committed nearly $10 billion to two larger, more complex follow-on tranches. DOD concurred with GAO's four recommendations but argued it was already implementing them; GAO disagreed. In January 2026, GAO reported the same pattern continuing: SDA had overestimated the technology readiness of critical hardware, still lacked an architecture-level schedule or a reliable life-cycle cost estimate, and combatant commands said they lacked insight into what capability they'd actually get or when.
Spend first, prove later
SDA builds PWSA in two-year phases it calls tranches: Tranche 0 was meant to demonstrate the architecture's core laser-communications technology before SDA committed to bigger, more complex follow-on tranches. As of December 2024, that demonstration hadn't gone well. Of Tranche 0's four prime contractors, the best-performing had demonstrated three of eight planned laser-communications capabilities; another had demonstrated one; the remaining two had demonstrated none. Despite that, SDA had already awarded Tranche 1 and Tranche 2 contracts worth a combined $9.9 billion -- about 15 times the $657 million spent on the Tranche 0 demonstration those follow-on tranches were supposed to build on -- for satellite constellations planned to be far larger and more technically complex than Tranche 0.
View data as table
| Tranche 0 (2020, the demo tranche) | 657 |
|---|---|
| Tranches 1 & 2 (2022-2024, awarded before T0 was fully demonstrated) | 9,941 |
View data as table
| Planned capabilities per T0 contractor | 8 |
|---|---|
| Demonstrated by the best-performing of T0's 4 contractors (Dec. 2024) | 3 |
DOD agreed, then argued nothing needed to change
's February 2025 report made four recommendations, chief among them that SDA demonstrate its minimum viable product for laser communications in Tranche 0 -- and later Tranche 1 -- before proceeding with launch decisions for the next tranche. The Department of Defense formally concurred with all four recommendations, with comments. But in those same comments, said it believed it was already implementing them. disagreed outright, writing that the evidence in its report showed SDA was not already taking the recommended actions. also noted that 's comments had proposed redefining Tranche 0's minimum viable product downward -- from demonstrating a working mesh network to merely demonstrating that building one was feasible.
Eleven months later: the same pattern, and a new one
's January 2026 report found SDA still at risk of falling behind its own schedule -- and, this time, overestimating how ready its technology actually was. SDA's internal documentation from August 2024 claimed that, before Tranche 1 and Tranche 2 began, its tracking satellites had already demonstrated specific technology readiness levels. found that wasn't accurate: one contractor was still in the process of demonstrating its Tranche 0 design in orbit, and none of the other three tracking-satellite contractors had demonstrated a tracking satellite on orbit before their tranches started.
No architecture-level schedule, no reliable cost estimate
also found SDA had continued awarding new tranche contracts every two years regardless of how earlier tranches were performing, without ever building a single schedule spanning the whole architecture -- SDA tracked each tranche's schedule separately. The Department of Defense still didn't know what the program would cost to build and sustain over time, because it lacked a reliable life-cycle cost estimate. And the combatant commands who will actually use the system told that SDA's requirements meetings functioned mainly as SDA presentations rather than a two-way process, leaving them unclear on when -- or whether -- their operational needs would be met.
Ground costs grew fast too
The satellites aren't the only thing that got more expensive. SDA's original Tranche 1 ground-control-segment contract -- covering two ground operations centers -- was awarded in May 2022 for $325 million. By July 2024, the contractor's own estimate to complete the work had grown to $445 million, a 37% increase. SDA separately added $491 million to the Tranche 2 ground segment through a sole-source contract modification. By May 2025, the Defense Contract Management Agency forecast the combined Tranche 1 and Tranche 2 ground segment cost would exceed $1 billion.
View data as table
| Original T1 ground segment contract (May 2022) | 325 |
|---|---|
| Contractor's estimate to complete (July 2024) | 445 |
's 2026 report made six new recommendations -- a technology readiness assessment for each future tranche's new hardware, real collaboration with combatant commands on requirements, a prioritized requirements backlog, an architecture-level schedule, contractor cost-data reporting, and a reliable, data-driven cost estimate. concurred with five of the six. It only partially concurred with the recommendation to require contractors to report actual cost data -- the same transparency gap had just finished describing as blocking a reliable cost estimate. said it continued to believe lacked the data it needed and left the recommendation unchanged.
The takeaway
- SDA kept spending before its technology was proven. It committed $9.9 billion to two bigger, more complex tranches -- 15 times the cost of the demonstration tranche -- while that demonstration tranche's best-performing contractor had shown only 3 of 8 planned capabilities.
- agreed with 's fix in words, not in practice. It concurred with all four 2025 recommendations while telling it was already doing what they asked; said the evidence showed otherwise, and flagged 's own attempt to lower the bar for what Tranche 0 had to prove.
- Eleven months later, found a new problem on top of the old one. SDA's claimed technology-readiness milestones for Tranche 1 and Tranche 2 weren't accurate, the architecture still has no unified schedule or reliable cost estimate, and only partially concurred with the recommendation to fix the cost-data gap behind that.
Findings on Tranche 0's unproven laser-communications technology, the $9.9 billion committed to Tranches 1 and 2 regardless, and 's 2025 response are from -25-106838, 'Laser Communications: Space Development Agency Should Create Links Between Development Phases' (February 26, 2025), read directly in full via an archived copy after the current gao.gov asset URL blocked direct access. Findings on SDA's overstated technology readiness, the missing architecture-level schedule and cost estimate, combatant-command transparency, ground-segment cost growth, and 's 2026 response are from -26-107085, 'Missile Warning Satellites: Space Development Agency Should Be More Realistic and Transparent About Risks to Capability Delivery' (January 28, 2026; revised February 2, 2026 to correct a word in one recommendation's text, with no change to substance), also read directly in full via an archived copy. The two reports examine the same satellite architecture about 11 months apart -- the first documenting unproven technology and continued spending, the second documenting overstated readiness claims and unresolved schedule and cost gaps.
Sources(2) ▾
- U.S. Government Accountability Office, Laser Communications: Space Development Agency Should Create Links Between Development Phases (2025-02-26) — -25-106838, a report to congressional committees finding that the Space Development Agency (SDA) had not yet demonstrated the core laser-communications technology of its Tranche 0 demonstration satellites, even as it awarded almost $10 billion in contracts for two larger, more complex follow-on tranches -- the technology-not-yet-proven facet. Distinct from doc-gao26-107085 (the risk-and-transparency facet, published about 11 months later, covering the same architecture's schedule, requirements, and cost problems). Direct gao.gov PDF fetch returns HTTP 403; fetched in full via the files.gao.gov HTML mirror (HTTP 200). gao.gov · original document
- U.S. Government Accountability Office, Missile Warning Satellites: Space Development Agency Should Be More Realistic and Transparent About Risks to Capability Delivery (2026-01-28) — -26-107085 (revised Feb. 2, 2026 to correct a word in Recommendation 3's text; no substantive change), a report to congressional committees examining the same PWSA architecture about 11 months after doc-gao25-106838, finding SDA overestimating technology readiness, lacking an architecture-level schedule and a reliable life-cycle cost estimate, and combatant commands reporting insufficient insight into requirements and delivery timing. Direct gao.gov PDF fetch returns HTTP 403; fetched in full via the files.gao.gov HTML mirror (HTTP 200). gao.gov · original document
Comments
Always open. Logged-in readers can annotate paragraphs in place.
The Space Development Agency's Proliferated Warfighter Space Architecture (PWSA) is a planned constellation of 300 to 500 satellites in low Earth orbit, split between a Transport Layer that moves data and a Tracking Layer that detects and tracks missile threats, including hypersonic weapons. In February 2025⧉, reported that SDA's demonstration tranche hadn't fully proven the laser-communications technology the whole architecture depends on -- and SDA had already committed billions more regardless. In January 2026⧉ -- about 11 months later -- reported on what came of that, and found new problems besides.