IBM and NASA open‑source a lunar AI model to turbocharge moon exploration

On September 10, 2026, IBM and NASA unveiled the NASA‑IBM Lunar Foundation Model, an open‑source AI system designed to accelerate scientific exploration of the Moon. The model is a large “foundation model” trained on vast amounts of lunar imagery and observational data, so researchers don’t have to build their own AI from scratch to run sophisticated analyses.

According to the announcement, the system can help with tasks such as mapping craters and boulder fields, assessing candidate landing zones, and identifying resources for potential future bases. Both the pretrained model and supporting code are being released openly, allowing universities and startups to use it at no cost—potentially democratizing access to high‑end “space AI” tools.

The move comes as U.S. and commercial lunar missions ramp up and the volume of raw data grows faster than human analysts can handle. Agencies and companies are increasingly turning to AI for automated interpretation of remote‑sensing data. IBM and NASA say they hope the lunar foundation model will become a shared platform for international collaboration, and that similar approaches could ultimately extend to Mars and other celestial bodies.