Astronomers used an AI-assisted tool called AnomalyMatch to analyze nearly 100 million images from NASA’s Hubble Space Telescope, identifying over 1,300 rare or unusual objects in just two and a half days, more than 800 of which had never been documented.

The anomalies included merging galaxies, gravitational lenses, star-forming clumps, jellyfish-like galaxies, and edge-on planet-forming disks, with some objects defying existing classification schemes.

Developed by ESA researchers David O’Ryan and Pablo Gómez, AnomalyMatch mimics human visual pattern recognition to efficiently flag potential anomalies, allowing astronomers to manually confirm the most promising findings.

This marks the first systematic search for astrophysical anomalies across the entire Hubble Legacy Archive and demonstrates the power of AI in handling the growing volume of astronomical data, NASA has reported.

The approach is expected to be crucial for future large-scale surveys from telescopes like the Nancy Grace Roman Space Telescope, Euclid, and the Vera C. Rubin Observatory.