NASA Administrator Jared Isaacman 宣布重组 Artemis 计划，以提高任务频率并削减成本。最重要的变动是取消 SLS 的 Exploration Upper Stage 与 Block 1B 升级；Artemis II 与 III 将继续使用现有上面级，而 Artemis IV、V 与后续任务将改用单一「标准化」上面级。时程也被重排：Artemis III 不再登月，而是在低地球轨道与 Starship 和／或 Blue Moon 对接；Artemis IV 现成为首次登月任务。NASA 目标自 2027 年中起每年发射 1 次 Artemis 任务，并在 2028 年至少完成 1 次登月。

这次调整的核心理由是 SLS 飞行速率过低。NASA 载人深空计划过去平均约每 3 个月发射 1 次，但自 Artemis I 以来已过近 3.5 年；官员并指出，Apollo 7 到 Apollo 8 仅相隔 9 周。NASA 在 Artemis I 与 II 的发射前准备中都遭遇氢气与氦气泄漏，导致以月计的延误。Jared Isaacman 认为，若继续把每枚火箭做成不同构型的「工艺品」，既无法提升可靠性，也无法与中国日益加快的月球计划竞争。新方案因此追求单一构型，并把发射节奏压缩到约每 10 个月 1 次。

新架构也改变了风险分配。NASA 不再让 Artemis III 直接从绕月飞越跳到完整登月，而是回到较接近 Apollo 7、8、9、10 的渐进式验证路线，以测试登月器性能、交会对接、通讯与太空衣等关键环节。不过，这会把压力转移到 Orion 的 2027 年中就绪、与商业登月器整合，以及新的商业上面级认证。另一个未解问题是 Lunar Gateway：Block 1B 被取消后，其用途更不明朗；相关第二发射塔成本已由 3.83 亿美元膨胀到 18 亿美元。Ars Technica 先前估计，若取消 Block 1B 与 Gateway，每年可释出约 20 亿美元。

NASA Administrator Jared Isaacman announced a restructuring of the Artemis program to raise mission cadence and cut costs. The biggest changes are cancellation of the SLS Exploration Upper Stage and Block 1B upgrade; Artemis II and III will keep the current upper stage, while Artemis IV, V, and later missions will shift to one standardized upper stage. The schedule also changes: Artemis III will no longer land on the moon, but instead dock in low Earth orbit with Starship and/or Blue Moon; Artemis IV now becomes the first lunar landing mission. NASA aims to begin flying 1 Artemis mission per year from mid-2027 and to achieve at least 1 lunar landing in 2028.

The main reason for the overhaul is the very low flight rate of SLS. NASA’s past human exploration programs launched crews about once every 3 months on average, but nearly 3.5 years have passed since Artemis I; officials also noted that only 9 weeks separated Apollo 7 and Apollo 8. NASA faced hydrogen and helium leaks during prelaunch preparations for both Artemis I and II, causing delays measured in months. Isaacman argues that continuing to treat each rocket as a custom configuration cannot improve reliability or keep pace with China’s accelerating lunar program. The new plan therefore favors a single configuration and a launch rhythm of about once every 10 months.

The new architecture also redistributes risk. Instead of asking Artemis III to jump directly from a crewed lunar flyby to a full landing, NASA is returning to a more Apollo-like progression, closer to Apollo 7, 8, 9, and 10, to test lander performance, rendezvous and docking, communications, and spacesuits. However, this shifts pressure onto Orion readiness for mid-2027, integration with commercial landers, and certification of a new commercial upper stage. Another unresolved issue is Lunar Gateway: with Block 1B canceled, its role is less clear, while the related second launch tower has grown from $383 million to $1.8 billion. Ars Technica had earlier estimated that canceling Block 1B and Gateway could free about $2 billion per year.