NASA’s Expanding Artemis Portfolio Tests Cost Discipline and Project Integration

In its July 2025 report to Congress, NASA: Assessments of Major Projects (GAO-25-107591), the U.S. Government Accountability Office (GAO) provides a revealing snapshot of NASA’s performance managing its largest and most ambitious space projects. The annual assessment focuses heavily on the agency’s Artemis program—an expansive suite of missions aimed at returning humans to the Moon and eventually reaching Mars. The report details both encouraging stability in near-term cost and schedule performance and growing structural risks tied to Artemis' complexity and interdependency.

This blog summarizes the findings of GAO report GAO-25-107591, authored by William Russell and the U.S. Government Accountability Office team. The full report is available at www.gao.gov/products/GAO-25-107591.

The GAO's 2025 analysis reviewed 38 major NASA projects with life-cycle costs exceeding $250 million. Among these, 18 are in active development and account for an estimated $74 billion in total life-cycle costs. Impressively, 14 of these projects experienced no cost or schedule growth since the 2024 report. However, the remaining four—most notably the Orion Multi-Purpose Crew Vehicle—collectively contributed over $500 million in cost overruns. Orion alone accounted for more than $360 million of that figure due to a combination of technical challenges, battery repairs, software issues, and post-Artemis I heatshield performance concerns.

While these annual fluctuations are concerning, the GAO points out that since it began tracking NASA’s major acquisitions in 2009, the majority of projects (30 of 53 that have completed or are nearing completion) have managed to stay under the statutory 15% cost overrun threshold that triggers mandatory congressional notification. That said, 12 projects did surpass the 30% threshold requiring reauthorization, including headline Artemis components like the Orion capsule and the Space Launch System (SLS) Block 1.

More than half of the cost overruns among all 53 assessed projects are attributable to just three Artemis programs: SLS Block 1, Orion, and the Exploration Ground Systems (EGS). Together, these three account for nearly $7 billion in overruns—an alarming signal that Artemis projects are fundamentally different in scope, scale, and technical risk than most prior NASA endeavors.

The GAO warns that the future may hold more of the same. NASA has initiated nine additional Artemis-related projects with estimated life-cycle costs exceeding $20 billion. These projects are deeply interdependent, and delays in one can cascade across the broader portfolio. Delays not only raise costs, they also reduce NASA’s agility in managing resources across missions. Indeed, the integration complexity of Artemis missions—a marked increase over previous programs—has become a critical variable in NASA’s ability to control costs.

To NASA’s credit, the agency has taken several steps to manage these growing risks. It created the Moon to Mars program office in 2023 to better coordinate Artemis activities, delayed setting cost baselines for several projects until requirements were more stable, and implemented new program management leadership roles. Despite these efforts, the GAO found that NASA had yet to fully implement two high-priority recommendations from earlier audits, including setting a separate cost and schedule baseline for Artemis II’s SLS Block I.

Perhaps most significantly, the GAO reiterates a critical lesson: underestimating technical complexity remains a recurring cause of project cost growth. As new Artemis missions advance from formulation into implementation, the likelihood of additional overruns remains high without disciplined cost estimation, firm requirements, and rigorous risk management practices.

NASA’s track record shows improvement in managing individual project performance, but the next wave of Artemis missions will test its ability to integrate, fund, and execute a system-of-systems architecture. The stakes—both financial and reputational—are as high as the ambitions of the Moon and Mars missions themselves.

Disclaimer: This blog post summarizes the findings of GAO-25-107591. While care has been taken to ensure accuracy, this post does not constitute legal, financial, or engineering advice. For official information, please consult the full GAO report.