Satellite runs a vision-language model to find targets on orbit

TechCrunch reported that a Loft Orbital YAM-9 satellite used NASA Jet Propulsion Laboratory software and Google DeepMind's Gemma 3 vision-language model to identify areas of interest in orbit without waiting for human analysts on the ground. The April demonstration is framed as the first reported use of a vision-language model in orbit. It matters because edge AI could let satellites triage sensor data before downlinking and eventually support always-on monitoring layers in space.

Key details: Published June 15, 2026, The demo ran on Loft Orbital's YAM-9 spacecraft, NASA JPL software used Google DeepMind's Gemma 3 vision-language model, The system responded to natural-language queries while in orbit.

Continue swiping for more AI Brief stories.