NASA's October 23 update now plans Artemis III as a low-Earth-orbit systems demonstration in 2027, with Artemis IV targeted as the first Artemis surface mission in 2028. These are announced plans, not completed milestones; schedules can change.
The case
GAO's July 2026 assessment covers 36 NASA major projects with at least $70 billion in estimated life-cycle costs. In February and March 2026, NASA announced significant changes to Artemis III, IV and V and paused work on three Artemis projects, including Gateway, while shifting more focus toward infrastructure for sustained lunar surface operations.
Major projects reported effects from NASA's 2025 civil-servant workforce reduction, according to GAO. Separately, GAO said Artemis architecture changes create acquisition-management challenges as projects adjust to new plans.
GAO reported that most projects in development had no annual schedule delays or cost overruns in the prior year, while portfolio cumulative cost growth and delays still increased slightly. Its concern was not that every project was failing. It was that major architecture and workforce changes create new management risks that old baselines may not capture.
The architecture-reset sequence
What mission outcome still survives the change?
Separate the enduring objective from the implementation architecture that was chosen earlier.
Which dependencies disappear, move or become newly critical?
Rebuild the interface map before reusing old schedules or readiness claims.
Which prior estimates are no longer decision-grade?
Preserve the old baseline for accountability, then create a clearly dated replacement rather than silently editing history.
Do the people and suppliers still match the changed architecture?
A program can retain headcount while losing the specific skill, facility, supplier or integration authority the revised mission needs.
What happens to paused, terminated or no-longer-central work?
Define closeout, data retention, hardware disposition, contract consequences and reusable assets.
What this changes
Assemble Teams analysis: architecture change should trigger a formal “mission reset,” not merely a schedule update. The most dangerous state is a new mission being managed with evidence, interfaces and capacity assumptions inherited from the old one.
What the evidence does not prove
GAO's report does not establish that NASA's revised Artemis architecture will succeed or fail, and this casebook does not assess the policy merits of the change. It uses the documented transition as a management case: when mission architecture changes materially, the evidence system must be re-baselined.
Use this after any major pivot.
Trigger the reset when a mission changes architecture, supplier, operating environment, funding mechanism or required outcome—not only when a schedule slips.
Browse Space & Aerospace missions →Editorial standard: this casebook describes documented NASA/GAO changes and labels Assemble Teams management synthesis separately.