Astronomers at the European Space Agency (ESA) have used artificial intelligence to uncover more than 1,400 previously undocumented astrophysical anomalies hidden within the Hubble Space Telescope’s 35-year archive. The discovery was led by researchers David O’Ryan and Pablo Gómez, who developed and deployed an AI model called AnomalyMatch to systematically scan through vast amounts of Hubble data. The model analyzed nearly 100 million image cutouts from the Hubble Legacy Archive—the first comprehensive, automated search of its kind. This massive dataset, while rich with information, presents a challenge: the sheer volume of data overwhelms traditional human-led analysis, and the universe is full of rare, unusual, or unexpected phenomena that can easily be overlooked. AnomalyMatch was trained to detect objects that deviate from typical patterns, flagging potential anomalies for further study. The AI completed the scan in just two and a half days—far faster than any human team could manage—identifying a wide range of unusual celestial features. Among the most common findings were galaxies in the process of merging or interacting, as well as gravitational lenses, where the light from distant objects is bent into arcs or rings by the gravity of massive foreground objects. Other notable anomalies included “jellyfish galaxies,” which exhibit long, trailing filaments of gas as they move through galaxy clusters, and galaxies with large, dense clumps of stars. Perhaps the most intriguing results were several dozen objects that defied classification entirely—strange, unexplained features that may represent entirely new types of astrophysical phenomena. “This is a fantastic use of AI to maximise the scientific output of the Hubble archive,” said Gómez. “Finding so many anomalous objects in Hubble data, where you might expect many to have already been found, is a great result. It also shows how useful this tool will be for other large datasets.” The findings, published in the journal Astronomy & Astrophysics, demonstrate the power of AI in unlocking new scientific insights from existing data. As space telescopes generate ever-larger datasets, such tools will become essential for extracting hidden knowledge from the cosmos. The success of AnomalyMatch suggests it could be adapted for use with data from other observatories, including the James Webb Space Telescope, opening new frontiers in astronomical discovery.