In a powerful reminder of how AI can accelerate science, astronomers have uncovered over 800 previously undocumented AI cosmic anomalies using a newly developed neural network—achieving in days what would have taken humans years.

The breakthrough comes from David O’Ryan and Pablo Gómez at the European Space Agency (ESA), who created an AI system called AnomalyMatch to scan astronomical images for rare and unusual objects. The results exceeded expectations by a wide margin.

How AnomalyMatch Found AI Cosmic Anomalies at Scale

AnomalyMatch was trained on the Hubble Legacy Archive, which contains tens of thousands of datasets collected over 35 years of Hubble observations. In just two and a half days, the AI processed nearly 100 million image cutouts, flagging 1,400 potential AI cosmic anomalies.

“While trained scientists excel at spotting cosmic anomalies, there’s simply too much Hubble data to analyze manually at this level of detail,” the ESA said in its announcement.

Human Verification Still Matters

Despite the speed and scale of the AI, human expertise remained essential. O’Ryan and Gómez manually reviewed the candidates returned by AnomalyMatch, confirming which objects were truly anomalous.

After verification, the astronomers confirmed that more than 800 of the AI cosmic anomalies had never been documented before, highlighting how much scientific value remains buried in existing datasets.

What the AI Cosmic Anomalies Revealed

Most of the newly identified AI cosmic anomalies fall into fascinating—but complex—categories:

• Interacting and merging galaxies, often with distorted shapes and extended tails of stars and gas
• Gravitational lenses, where massive foreground galaxies warp spacetime and bend light into arcs or rings
• Edge-on planet-forming disks, offering rare views into early planetary systems
• Jellyfish galaxies and galaxies with unusually large stellar clumps

Adding intrigue, the researchers also identified dozens of objects that defy current classification, suggesting there may be cosmic phenomena still poorly understood—or entirely new.

Why This Matters for Astronomy

“This is a fantastic use of AI to maximize the scientific output of the Hubble archive,” Gómez said in ESA’s release. “Finding so many anomalous objects in data where you’d expect most discoveries to already be made is a remarkable result.”

Beyond Hubble, the success of AnomalyMatch suggests AI could play a central role in analyzing future mega-datasets from upcoming observatories like the James Webb Space Telescope and ESA’s Euclid mission.

In short, AI cosmic anomalies are proving that even well-studied data can still hide groundbreaking discoveries—if you know how to look.

Don’t miss out on our latest news—follow us for the latest AI news, breakthroughs, and insights that matter.