Generative AI Gives Spacecraft the Autonomy Engineers Once Feared
TL;DR
NASA and its partners are testing generative AI to give spacecraft more autonomy. Last December, NASA's Jet Propulsion Laboratory used Anthropic's Claude models to help plan two Mars drives for the Perseverance rover, with human planners checking and adjusting the route before upload. In May, NASA and IBM ran a compressed AI model on the International Space Station and a satellite to identify floods and clouds from orbit, the first model of its kind demonstrated in space.
Nauti's Take
The first useful test for a small AI team is a tightly bounded offline workflow with human approval: which decisions may the model prepare, and when must a fallback take over? Before production use, verify audit logs, data quality, model versions, and behavior during network loss.
These space experiments show serious interest in autonomy, not operational readiness.
Sources