NASA has revealed that its Perseverance rover successfully completed the first-ever AI-planned drive on another planet, using Anthropic’s Claude to assist with navigation planning on Mars. According to Anthropic, the AI helped map a roughly 400-meter route that the six-wheeled rover navigated across the Martian surface in December.

Engineers trained Claude on years of rover driving data, after which the AI generated navigation commands and plotted waypoints across complex terrain. Claude analyzed orbital imagery to chart a safe path through rocks and sand ripples, assembling a route that it then self-critiqued and refined.

Before transmitting any commands to Mars, NASA engineers validated Claude’s proposed routes using simulation modeling. The final plan required only minor human adjustments, demonstrating that AI-assisted planning can integrate smoothly into existing mission safety workflows.

NASA engineers said the approach could cut route-mapping time by up to 50%, allowing operators to schedule more drives and collect additional scientific data during the rover’s limited operational windows.

Why it matters: AI has moved far beyond office productivity tools and software development — it is now helping navigate autonomous robots on another planet. If systems like Claude can reliably assist with mission-critical planning 140 million miles from Earth, the scope of tasks where AI can meaningfully augment human decision-making is expanding faster than most expected.