Artificial intelligence is helping astronomers uncover hidden cosmic phenomena across billions of celestial objects. In a major discovery, an international team of scientists has confirmed 70 new gravitational lenses, offering fresh opportunities to investigate dark matter, distant galaxies, and the evolution of the universe.
The universe contains many mysteries that scientists are still trying to understand. Among them are dark matter, the nature of distant galaxies, and the forces shaping the cosmos over billions of years. Now, researchers have discovered 70 new gravitational lenses using artificial intelligence (AI), powerful telescopes, and one of the largest astronomical imaging datasets ever assembled.
These rare cosmic structures act like natural magnifying glasses, bending and magnifying light from distant objects. By studying them, astronomers can observe galaxies that might otherwise be too faint or distant to detect.
The discovery expands one of the largest collections of confirmed gravitational lenses and provides scientists with valuable new tools for exploring the universe.
A Massive Map of the Universe Opens New Possibilities
The discovery builds on the DESI Legacy Imaging Surveys, an ambitious astronomical project designed to capture detailed images of the sky.
In August 2026, the survey team released an extraordinary map containing approximately 5.6 trillion pixels and nearly four billion celestial objects. These include stars, galaxies, black holes, and asteroids.
Creating such a detailed map required 13 years of observations and data processing. The project combined information from three major sky surveys.
The Dark Energy Camera Legacy Survey used the powerful Dark Energy Camera mounted on the Víctor M. Blanco 4-meter Telescope in Chile.
The Mayall z-band Legacy Survey collected observations using the Nicholas U. Mayall 4-meter Telescope in Arizona.
The Beijing-Arizona Sky Survey combined observations from the Bok 2.3-meter Telescope in Arizona with additional infrared data from NASA's Wide-field Infrared Survey Explorer satellite.
Together, these surveys created an enormous collection of astronomical images. Although the primary observations were designed to support studies of the universe's large-scale structure, the dataset has also become a valuable resource for discovering unusual objects.
Gravitational lenses are among the most exciting examples.
What Are Gravitational Lenses?
Gravitational lensing occurs when a massive object lies between Earth and a more distant object in space.
According to Einstein's theory of general relativity, gravity can bend the path of light. When light from a distant galaxy passes near a massive foreground galaxy or galaxy cluster, the foreground object's gravity changes the light's path.
Under the right conditions, the background galaxy can appear brighter, stretched into curved arcs, or duplicated into several images. Sometimes, the light forms an almost complete circle called an Einstein ring.
The foreground object acts like a natural telescope, allowing astronomers to examine distant cosmic objects with greater detail than would otherwise be possible.
However, these alignments are rare. The background galaxy, foreground object, and observer must be positioned in a suitable arrangement for a strong lensing effect to occur.
Finding these systems among billions of astronomical objects is therefore a major challenge.
How Artificial Intelligence Found Hidden Cosmic Structures
To search for gravitational lenses efficiently, the research team turned to artificial intelligence.
Led by Xiaosheng Huang of Santa Clara University and the U.S. Department of Energy's Lawrence Berkeley National Laboratory, researchers used machine-learning techniques to examine the vast Legacy Surveys dataset.
Their approach relied on a type of AI called a residual neural network. This system can learn to recognize complex visual patterns in images, including the curved arcs and unusual shapes associated with gravitational lensing.
Unlike a traditional search that depends on astronomers manually inspecting images one by one, an AI system can examine enormous numbers of images and identify those most likely to contain interesting objects.
Earlier work using these methods had already produced a catalog of approximately 3,500 potential gravitational lenses.
The latest search identified 76 promising candidates for further investigation.
However, AI can only identify objects that resemble gravitational lenses. It cannot independently establish that every candidate is a genuine lens. Some ordinary galaxies and other astronomical structures can produce similar visual patterns.
The researchers therefore needed additional observations to verify the candidates.
Powerful Telescope Observations Confirm 70 Discoveries
To determine which candidates were genuine gravitational lenses, the team conducted follow-up observations using the Multi-Unit Spectroscopic Explorer (MUSE), an advanced instrument installed on the European Southern Observatory's Very Large Telescope.
The observations were collected between 2022 and 2024.
MUSE uses a technique known as integral field spectroscopy. Rather than collecting a single spectrum from a small region, it records spectral information across an entire observed field.
This produces a dataset containing both spatial information and information about the wavelengths of light coming from different parts of an image.
Why is this important?
Light from distant galaxies carries clues about their movement, chemical composition, and distance from Earth. By examining these clues, astronomers can distinguish the foreground galaxy producing the lensing effect from the more distant background object.
This step is essential because a genuine gravitational lens involves two separate objects positioned at different distances, even if they appear close together in an image.
After analyzing the spectroscopic observations, the researchers confirmed that 70 of the 76 candidates were genuine gravitational lenses.
The result demonstrates how AI and advanced telescope observations can work together: machine learning identifies promising targets, while spectroscopy provides the evidence needed to confirm them.
How These Discoveries Could Help Solve the Dark Matter Mystery
One of the most important applications of gravitational lensing is the study of dark matter.
Dark matter is an invisible form of matter that scientists believe plays a major role in the formation and evolution of galaxies. It does not emit or reflect light in the way ordinary matter does, making it difficult to observe directly.
Yet dark matter has gravity, and its gravitational influence can affect the paths of light traveling through space.
By studying how a background galaxy's light bends, astronomers can estimate the distribution of mass in the foreground system. Comparing that distribution with the visible stars and gas helps researchers investigate where dark matter may be located.
A larger collection of confirmed gravitational lenses gives scientists more opportunities to test their models of how dark matter is distributed throughout the universe.
These systems can also reveal details about galaxy structure, including how matter is arranged within galaxies and how galaxies develop over cosmic time.
A Window Into the Distant Universe
Gravitational lenses can also help scientists study galaxies from the early universe.
Because light takes time to travel across space, observing extremely distant galaxies means observing them as they appeared billions of years ago.
Some of these galaxies are so faint that ordinary telescopes struggle to collect enough light to study them in detail. Gravitational lensing can increase their apparent brightness and stretch their images, making certain features easier to investigate.
This can help astronomers examine early star formation, galaxy development, and the physical conditions that existed when the universe was much younger.
The new discoveries may become particularly valuable when combined with observations from major space telescopes and upcoming or ongoing astronomical surveys, including the Hubble Space Telescope, the James Webb Space Telescope, the Nancy Grace Roman Space Telescope, and the Vera C. Rubin Observatory's Legacy Survey of Space and Time.
Together, these resources can help scientists build a more detailed picture of the universe's history.
Gravitational lensing can also contribute to measurements of the universe's expansion rate and, in suitable systems, support investigations of planets orbiting other stars. However, these applications require specific observational conditions and additional analysis.
Building a Valuable Resource for Future Astronomy
The 70 newly confirmed lenses have been added to the DESI Strong Lens Foundry Project, expanding a major collection of verified gravitational lens systems.
According to Aleksandar Cikota, an associate scientist at NSF NOIRLab and a co-author of the research, deep and wide-field images from the Legacy Surveys have created an unprecedented foundation for finding these rare objects.
The discovery also highlights a broader change in astronomy. Modern researchers increasingly combine enormous datasets, AI algorithms, high-performance computing, and advanced telescopes to investigate questions that were once difficult to approach.
Computing resources from the National Energy Research Scientific Computing Center also supported the broader search effort, showing how computational infrastructure has become an important part of astronomical research.
Although these 70 discoveries will not solve every mystery about dark matter or cosmic evolution on their own, they give scientists more systems to investigate and compare.
Each confirmed lens adds another opportunity to study how gravity shapes the universe, how galaxies develop, and how invisible matter influences the cosmos.
As astronomical surveys continue to collect enormous amounts of data, AI-assisted searches could uncover many more hidden cosmic structures.
The universe may contain billions of objects that remain difficult to identify with traditional methods. By combining artificial intelligence with powerful telescopes, scientists are finding new ways to uncover these hidden structures—and bringing some of the universe's deepest mysteries within reach.
Research reference: Emerald Lin et al., “DESI Strong Lens Foundry. IV. Spectroscopic Confirmation of DESI Lens Candidates with VLT/MUSE,” The Astrophysical Journal Supplement Series (2026). DOI: 10.3847/1538-4365/ae844d.
Comments
Post a Comment