NASA asks volunteers to help AI distinguish telescope errors from galaxies.

NASA introduced Artifact InSPECtor, a citizen-science project in which volunteers identify unwanted signals in telescope data. Reflections, detector effects and cosmic rays can create features that software mistakes for astronomical objects. Participants check the AI's classifications to help improve the system.

The project initially uses observations from ESA's Euclid mission, which includes NASA contributions. NASA says data from its Nancy Grace Roman Space Telescope is expected to join the project in early 2027. Better-cleaned spectra can support measurements of galaxies and research into the universe's expansion; the project is not itself a new discovery about dark energy.