NASA Plotted Perseverance Rover's Route on Mars Using Claude AI

A Historic Collaboration: AI Guides the Way on Mars

NASA's Perseverance rover has passed a significant AI-assisted milestone in its mission on Mars. In December 2025, the vehicle successfully followed a route planned by Anthropic's Claude large language model in the Jezero Crater. This marks the first time NASA has used an AI model to plan the journey of an interplanetary robot.

Optimizing a Complex Task with AI

Charting a course for Perseverance is a laborious and meticulous process, even for human operators. Every movement must be carefully pre-planned to prevent the rover from slipping, tipping over, or having its wheels spin in place. Using the traditional method, operators create a 'track route' using images from space and data from the rover's cameras.

While Claude Code (Anthropic's programming agent) was trained for this task, it had to be fed with years of contextual data from the exploration vehicle. The model then systematically assembled waypoints consisting of ten-meter segments, which it would later critique and rework.

Human Oversight and Efficiency Gains

Engineers at NASA's Jet Propulsion Laboratory (JPL) double-checked the AI's work by running it through a simulation they use to verify commands before sending them to the rover each day. NASA reported having to make 'minor adjustments' to the route, including one change made because the team had access to ground-level images that Claude did not see during its planning process.

According to engineers' estimates, using Claude in this way will cut route planning time in half and make journeys more consistent. Operators spending less time on manual planning and training will enable more driving, more scientific data collection, and more analysis.

Pointing to the Future of Space Exploration

NASA expressed its excitement that autonomous AI systems could help probes explore more distant regions of the solar system. This collaboration is seen as a significant step demonstrating the potential of AI models in complex, real-world missions.

This development constitutes another example of how AI's applications in the physical world are rapidly expanding, similar to companies shifting production lines to humanoid robots. For NASA, any tool that enables more efficient work under budget and personnel constraints is of critical importance.