NASA's Mars Rover Is Now Driven by Claude: AI Takes the Wheel on the Red Planet
By Sanna the Weaver • Sun Mar 01 2026 • Technology
On a cold Martian morning in February 2026, the Perseverance rover drove 340 meters across the Jezero Crater floor along a path it had never been shown by a human engineer. The route had been planned entirely by an artificial intelligence system — Anthropic's Claude vision-language model — which analyzed orbital imagery and terrain data, identified safe waypoints, and generated a drive plan that mission controllers reviewed and approved before execution. It was the first time in the history of planetary exploration that an AI system had autonomously planned a rover drive on another world. Why AI Planning Matters for Mars The fundamental challenge of operating a rover on Mars is the communication delay. At current orbital positions, a signal takes between 3 and 22 minutes to travel between Earth and Mars — meaning that real-time remote control is impossible, and every rover drive requires a plan transmitted hours or days in advance. NASA's rover operators currently spend substantial time analyzing terrain images and manually planning safe routes, a process that limits how far Perseverance can travel in any given Mars day. AI planning could dramatically accelerate exploration by automating this process. How Claude Was Adapted Claude was not designed for Mars rover operations, but its vision-language capabilities — understanding both images and spatial instructions — made it a natural candidate for terrain analysis tasks. NASA's Jet Propulsion Laboratory worked with Anthropic to fine-tune Claude on thousands of Martian terrain images, teaching it to identify hazards (loose rocks, steep slopes, soft sandy terrain) and optimal paths. The model generates not just a recommended route but a confidence score for each waypoint and an explanation of its reasoning — allowing human controllers to spot-check the AI's logic before approving any drive. "We are not taking humans out of the loop. We are making the loop faster and more informed." — NASA JPL Mission Director, February 2026 Implications for Future Missions The success of AI-planned Perseverance drives has immediate implications for NASA's future Mars missions. The proposed Mars Sample Return mission — which would require a lander to autonomously navigate to a cache of samples collected by Perseverance and then launch them into Mars orbit — will depend heavily on autonomous AI decision-making in scenarios where human communication delays make real-time control impossible. Further ahead, any crewed Mars mission would require AI systems capable of autonomous decision-making for operations ranging from surface navigation to medical emergencies. The Perseverance experiment is a proof of concept for AI as an indispensable partner in deep space exploration.