- \n
- AI sifted through nearly 100 million Hubble image cutouts<\/strong>, identifying over 1,300 unusual objects in just a few days \u2014 more than 800 of them never before documented in scientific literature.<\/li>\n
- Most of the anomalies are rare or odd<\/strong> cosmic phenomena \u2014 including merging galaxies, gravitational lenses, ring galaxies, and galaxies with unusual shapes or features.<\/li>\n
- A new AI tool called AnomalyMatch made the discoveries possible,<\/strong> allowing astronomers to rapidly comb the vast Hubble archive and highlight objects that would be extremely time-consuming to find manually.<\/li>\n<\/ul>\n
NASA originally published this article on January 27, 2026. Read the original here. Edits by EarthSky.<\/p>\n
AI combed Hubble\u2019s archive<\/h3>\n
A team of astronomers has employed a cutting-edge, artificial intelligence-assisted technique to uncover rare astronomical phenomena within archived data from the NASA-ESA Hubble Space Telescope. The team analyzed nearly 100 million image cutouts from the Hubble Legacy Archive. These are small, cropped sections of much larger images, each measuring just a few dozen pixels (7 to 8 arcseconds) on a side. They identified more than 1,300 objects with an odd appearance in just two-and-a-half days \u2014 more than 800 of which had never been documented in scientific literature.<\/p>\n
Most of the anomalies were galaxies undergoing mergers or interactions, which exhibit unusual morphologies or trailing, elongated streams of stars and gas. Others were gravitational lenses, where the gravity of a foreground galaxy distorts spacetime and bends light from a background galaxy into arcs or rings. <\/p>\n
Additional discoveries included galaxies with massive star-forming clumps, jellyfish-looking galaxies with gaseous \u201ctentacles,\u201d and edge-on planet-forming disks in our own galaxy resembling hamburgers. <\/p>\n
Remarkably, several dozen objects defied existing classification schemes entirely!<\/p>\n
![\"Young](\"https:\/\/earthsky.org\/upl\/2026\/01\/david-patrick-oryan-ESA-e1769537964552.jpg\")
This is David Patrick O\u2019Ryan, lead author of the new paper describing how AI combed Hubble\u2019s archive. O\u2019Ryan wrote, \u201cThe main focus of my research remains on the relationship between galaxy evolution and galaxy morphology. Using machine learning algorithms, we can quickly get morphology classifications of many tens to hundreds of thousands of galaxies. This allows us to statistically investigate the relationship between different morphologies and physical processes of the universe.\u201d Image via ESA.<\/figcaption><\/figure>\n EarthSky\u2019s 2026 lunar calendar is available now. Get yours today! Makes a great gift.<\/p>\n
The formidable challenge of identifying types of galaxies<\/h3>\n
Identifying such a diverse array of rare objects within the vast and growing repository of Hubble and other telescope data presents a formidable challenge. Never in the history of astronomy has such a volume of observational data been available for analysis.<\/p>\n
To address this challenge, researchers David O\u2019Ryan and Pablo G\u00f3mez of ESA (the European Space Agency) developed an AI tool capable of inspecting millions of astronomical images in a fraction of the time required by human experts. Their neural network, named AnomalyMatch, was trained to detect rare and unusual objects by recognizing patterns in data \u2014 mimicking the way the human brain processes visual information. David O\u2019Ryan, lead author of the study published in Astronomy & Astrophysics, said:<\/p>\n
\n
Archival observations from the Hubble Space Telescope now span 35 years, offering a rich dataset in which astrophysical anomalies may be hidden.<\/p>\n<\/blockquote>\n
Traditionally, anomalous images are discovered through manual inspection or serendipitous observation. While expert astronomers excel at identifying unusual features, the sheer volume of Hubble data makes comprehensive manual review impractical. Citizen science initiatives have helped expand the scope of data analysis. But even these efforts fall short when faced with archives as extensive as Hubble\u2019s or those from wide-field survey telescopes like Euclid, an ESA mission with NASA contributions.<\/p>\n
This is a significant advancement<\/h3>\n
The work by O\u2019Ryan and G\u00f3mez represents a significant advancement. By applying AnomalyMatch to the Hubble Legacy Archive, they conducted the first systematic search for astrophysical anomalies across the entire dataset. After the algorithm flagged likely candidates, the researchers manually reviewed the top-rated sources and confirmed more than 1,300 as true anomalies. G\u00f3mez commented:<\/p>\n
\n
This is a powerful demonstration of how AI can enhance the scientific return of archival datasets. The discovery of so many previously undocumented anomalies in Hubble data underscores the tool\u2019s potential for future surveys.<\/p>\n<\/blockquote>\n
Bottom line: Guided by astronomers, AI combed Hubble\u2019s archive \u2013 some 100 million Hubble image cutouts \u2013 and identified more than 1,300 unusual objects.<\/p>\n
Source: Identifying astrophysical anomalies in 99.6 million source cutouts from the Hubble legacy archive using AnomalyMatch<\/p>\n
Via NASA<\/p>\n<\/div>\n
\n - Most of the anomalies are rare or odd<\/strong> cosmic phenomena \u2014 including merging galaxies, gravitational lenses, ring galaxies, and galaxies with unusual shapes or features.<\/li>\n