Three days before NASA's Flight Readiness Review cleared Artemis II for an April 1 launch, the agency's own watchdog published a report that deserves more attention than it has received. The NASA Office of Inspector General released audit IG-26-004 on March 10, examining the Human Landing System program — and what it found was sobering. Both SpaceX and Blue Origin are behind schedule, and NASA currently has no capability to rescue astronauts stranded in space or on the lunar surface if something goes wrong. The report covers a program that has grown to nearly billion in obligated contracts since 2019, with projections reaching over 8 billion through fiscal year 2030. NASA has controlled costs reasonably well — SpaceX's contract grew just 6 percent, Blue Origin's less than 1 percent — but the OIG found gaps in testing and crew survival planning that the agency needs to close before any astronaut descends to the Moon. AI-generated image NASA's OIG released report IG-26-004 on March 10, 2026, flagging unmitigated safety risks in the Human Landing System program. Credit: AI illustration Nearly Billion Spent, More Than 8 Billion to Go NASA's Human Landing System program is one of the most complex and expensive commercial partnerships in agency history. The basic structure: NASA picks contractors, funds their development, embeds government experts in their facilities, and holds them accountable to requirements — but gives them wide latitude in how they get there. The OIG assessed that approach as generally sound, with the collaborative model helping both providers navigate technical hurdles faster than a traditional cost-plus arrangement might allow. B Obligated to HLS since 2019 8B+ Projected through FY 2030 +6% SpaceX contract growth <1% Blue Origin contract growth 2028 Current target for first crewed landing 5 OIG recommendations issued The cost discipline is real and worth acknowledging. In federal space procurement, a 6 percent contract growth for a system as novel as Starship HLS is genuinely unusual. But cost control is only part of the picture. The OIG found that both providers have experienced schedule delays and technical difficulties that the agency needs to manage more aggressively — and that the human safety framework has holes in it that government program managers cannot afford to leave open. Starship HLS: The Elevator Problem and Other Risks SpaceX is building a purpose-built variant of Starship for the Artemis III and IV crewed lunar landings, currently targeting 2028 after schedule slips pushed out the original 2027 window. The OIG flagged several specific technical risks that NASA and SpaceX are actively working through, some with more urgency than others. AI-generated image Starship HLS stands roughly 171 feet tall on the lunar surface. The crew elevator descends from approximately 115 feet, with no backup ingress system currently designed. Credit: AI illustration The most striking risk in the report involves height. Starship HLS is roughly 171 feet tall — about the size of a 14-story building — when standing on the lunar surface. At the south pole, where NASA plans to land for the water-ice resources and extended solar access, the terrain includes slopes up to 20 degrees and boulders exceeding 65 feet in height. A tip-over scenario on that kind of terrain would be catastrophic, and the OIG notes it as an unresolved risk in lander development planning. The elevator situation is arguably the more immediate concern. Crew access to and from the lunar surface requires descending roughly 115 feet via a single elevator mechanism. The OIG identified this as a top risk category — and critically, there is currently no backup ingress system if the elevator fails. SpaceX and NASA's HLS program office are working on alternative access concepts, but none are finalized. A crew stranded 115 feet above the surface, unable to get down or get back to the vehicle, faces a scenario the current safety architecture does not address. Top Starship HLS Loss-of-Crew Risk Contributors • Avionics: Autonomous descent and landing systems must function without real-time Earth intervention due to signal delays. • Main engines: Raptor Vacuum variants must perform reliably in the lunar thermal and vacuum environment. • Propulsion/cryogenics: Long-duration cryo propellant storage in cislunar space remains a technical challenge under development. • Landing legs: Structural requirements on uneven regolith at a 171-ft-tall vehicle create tip risk in ways Apollo-scale landers never faced. • Electrical power: A power failure during surface operations could cascade into life support and communication failures. The report also highlights a disagreement between NASA and SpaceX over manual control during descent and landing abort scenarios. Human-rating standards require single-failure tolerance — meaning a crew must be able to abort even if one system fails. The OIG found that SpaceX's current design approach may not meet this standard without a waiver, and that NASA has not yet determined whether a waiver is appropriate. The Commercial Crew program, which certified Dragon and Starliner, went through extensive manual control analysis that the OIG recommends applying to Starship HLS before any crew boards. On schedule: SpaceX completed its first orbital cryogenic propellant transfer test in early 2026, running about 12 months behind original milestones. The critical design review is now targeting August 2026. An uncrewed landing demonstration is planned for late 2026. The crewed Artemis III mission targets no earlier than 2028 — and that timeline requires every remaining milestone to hit on schedule, which has not been a consistent pattern for this program. Blue Moon: Terrain Challenges and Immature Cryo Tech Blue Origin's Blue Moon Mark 2 lander is assigned to Artemis V and beyond, targeting the mid-2030 timeframe. Its development profile is earlier in the cycle than Starship HLS — which means some technical questions that are becoming urgent for SpaceX are still in earlier design phases for Blue Origin. The OIG flagged that as a risk in itself: if technical challenges surface late, there is less runway to resolve them before the Artemis V mission window. AI-generated image SpaceX Starship HLS (left) and Blue Origin Blue Moon (right) represent two fundamentally different architectural approaches to the same problem. Credit: AI illustration Blue Moon's geometry creates a different set of terrain risks than Starship. The lander has a lower center of gravity and wider footprint, but the OIG found that tilts greater than 8 degrees exceed the vehicle's design tolerance — a problem on south pole terrain where permanently shadowed craters can be as deep as the Grand Canyon and slopes in the target landing zones vary significantly. If the landing site assessment is wrong, or if the terrain is disturbed during descent, the vehicle could settle at an angle that prevents hatch operations or safe crew egress. The manual control question applies to Blue Moon as well. As of November 2025, Blue Origin had not finalized key design decisions on manual control capability. Human-rating requirements are clear: crews must be able to intervene manually, and the system must tolerate single failures. Blue Origin has not yet sought or received a waiver on this requirement, meaning the design work is not complete. Metric SpaceX Starship HLS Blue Origin Blue Moon Assigned Mission Artemis III, IV Artemis V+ Target Landing Date ~2028 ~March 2030 CDR Schedule Slip ~12 months ~11 months Uncrewed Demo Slip ~12 months (end-2026) ~8 months (Feb 2029) Key Open Risk Elevator backup, manual control waiver Cryo maturity, tilt tolerance on terrain Contract Cost Growth +6% <1% Cryogenic propellant management is Blue Origin's other identified technical gap. Long-duration storage of liquid hydrogen and liquid oxygen in cislunar space involves managing boil-off, thermal gradients, and pressurization with hardware that