Dyson Swarm Candidates: What the Latest Infrared Star Surveys Actually Found

 

The idea of a Dyson swarm has a wonderfully sci-fi ring to it: a vast cloud of solar-collecting structures orbiting a star, capturing its energy and radiating the leftover heat into space. It sounds like something dreamed up for a space opera—or an especially ambitious homeowner’s association.

But it’s also a legitimate scientific idea. Physicist Freeman Dyson suggested in 1960 that a technologically advanced civilization might eventually use so much energy that it would need to surround its star with energy-harvesting infrastructure. Not a solid shell—that would be mechanically absurd—but potentially a sprawling swarm of satellites, habitats, mirrors, and collectors.

If such a system existed, astronomers would not necessarily see little alien panels glittering in the darkness. Instead, they might see something more mundane and more measurable: infrared excess.

That’s because any machine using stellar energy must eventually dump waste heat. A Dyson swarm could intercept visible light from its star and re-emit part of that energy as infrared radiation, making the system appear unexpectedly bright at mid-infrared wavelengths.

Recently, large surveys combining data from Gaia, 2MASS, and WISE have produced a small but fascinating set of stars with unusual infrared signatures. None is proof of aliens. Not remotely. But they are precisely the kind of oddities worth examining carefully.

Here are the most important recent results—and why the story is both more interesting and more cautious than the headlines suggest.

1. Project Hephaistos searched millions of nearby stars for infrared waste heat

The most prominent modern search is Project Hephaistos, a research program designed to look for possible Dyson-swarm-like signatures in enormous astronomical databases.

Its 2024 study combined observations from:

• Gaia DR3, which provides exceptionally precise distances, brightness measurements, and stellar properties;
• 2MASS, a near-infrared all-sky survey;
• WISE, the Wide-field Infrared Survey Explorer, which mapped the sky in mid-infrared light.

The team examined roughly five million stars within 300 parsecs—about 978 light-years—of Earth. That’s a huge sample. And it matters, because Dyson swarm searches are not really about finding one weird star and getting excited. They are about looking systematically enough that the weirdness has statistical meaning.

The researchers asked a simple question: does a star emit more infrared radiation than its temperature, size, and distance predict?

Most stars have well-understood spectral energy distributions. A cool red dwarf should glow in a particular way. A Sun-like star has its own expected pattern. If the observed infrared light greatly exceeds the expected amount, something is adding extra heat or light to the system.

That “something” could be a Dyson swarm. It could also be dust, a background galaxy, an active galactic nucleus, a bad measurement, or an unlucky overlap of unrelated objects in the same patch of sky.

Astronomy, as ever, is the art of finding a cosmic mystery and then spending years discovering it was dust. Still, dust is not always the answer—and the search is valuable either way.

The Project Hephaistos team used automated filtering, including machine-learning-assisted checks and visual inspection, to reduce the chances that WISE’s relatively broad infrared images were being confused by nearby sources. Their work was published in the Monthly Notices of the Royal Astronomical Society as Project Hephaistos II.

2. Seven red dwarfs stood out as especially unusual candidates

The 2024 Project Hephaistos analysis identified seven M-dwarf stars with strong mid-infrared excesses that initially appeared consistent with partial Dyson swarms.

M dwarfs, or red dwarfs, are the smallest and most common stars in the Milky Way. They are cooler and dimmer than the Sun, but they have one enormous advantage for any hypothetical civilization: they can remain stable for extraordinarily long periods. Some could shine for trillions of years.

From an engineering perspective, a swarm around a red dwarf could also be more compact. A civilization would need to build its collectors closer to the star to receive enough energy, reducing the total area that needs to be covered. “Easier” is doing some heroic lifting in that sentence, of course. Building a star-scale power system remains a fairly ambitious weekend project.

The candidates’ infrared signatures were striking. In some cases, the systems appeared much brighter in infrared light than expected from the visible starlight alone—up to dozens of times brighter in the relevant measurements.

The initial modeling suggested that, if the excess really originated around the stars, it could be compatible with structures intercepting a significant fraction of the stars’ energy and radiating waste heat at temperatures broadly in the range expected for warm engineered surfaces.

That last phrase is important: compatible with is not the same as evidence for.

A partial Dyson swarm would not block every ray of starlight. The star could still be visible in optical data while a fraction of its light was captured, used, and eventually re-emitted as infrared heat. This is exactly the sort of spectral mismatch the survey was designed to find.

Yet infrared excess is a clue, not a conclusion.

3. The candidates are intriguing because ordinary explanations are not always obvious

Why did these seven stars attract attention? Because the research team had already tried to account for several standard explanations.

Young stars, for example, can be surrounded by dusty disks left over from planet formation. These disks absorb starlight and glow in infrared wavelengths. Older stars can have debris disks created by asteroid collisions, comet activity, or the long-term grinding of planetary material.

But the Hephaistos candidates were not straightforward examples of those familiar systems. Their infrared profiles did not neatly resemble ordinary dusty disks, and some appeared unusually extreme for their stellar types.

That does not mean the objects are artificial. It means they belong in the “needs more observation” folder—the most scientifically honorable folder in astronomy.

A genuine Dyson swarm candidate should ideally satisfy several conditions at once:

1. The infrared source should be physically associated with the star, not a distant object sitting along the same line of sight.
2. Its spectrum should resemble smooth thermal waste heat, rather than the chemically distinctive emission of ordinary dust.
3. The system should not show evidence of a conventional debris disk, stellar companion, galaxy, or active black hole.
4. The signal should persist across independent observations and withstand better-resolution follow-up.
5. The star’s overall energy budget should make sense. If huge amounts of starlight are intercepted, astronomers need to understand where that missing visible energy went.

That is a high bar. It should be. Claiming evidence of technology beyond Earth ought to require more than a funny-looking infrared curve.

4. JWST follow-up has already ruled out some of the most tempting possibilities

Here is where the story takes a turn that is less sensational but much more scientific.

In July 2026, a follow-up study using the James Webb Space Telescope examined two Project Hephaistos candidates with far better infrared resolution than WISE could provide. The results strongly indicated that the infrared excesses did not come from structures around the candidate stars.

Instead, Webb resolved nearby background galaxies that had blended with the stars in the lower-resolution survey data.

One case appeared to involve a heavily dust-obscured galaxy, while another involved a dusty star-forming galaxy. In both instances, the background object sat extremely close to the star in the sky—close enough that WISE had difficulty separating their infrared light.

That is not a failure of the original survey. It is precisely how science is supposed to work.

Wide-field surveys like WISE are brilliant at identifying unusual targets across the entire sky. But their images are comparatively coarse. If a distant infrared-bright galaxy happens to lie almost directly behind or beside a nearby star, the two sources can merge into what looks like a single strange object.

JWST, with its vastly sharper infrared vision, can untangle those overlaps. The relevant follow-up work is available as an arXiv preprint.

This result narrows the field. It does not eliminate the value of Dyson swarm searches. In fact, it teaches researchers exactly what future searches need: higher-resolution confirmation, multiwavelength observations, and a healthy suspicion of cosmic photobombers.

5. Several other candidates may also have natural explanations

The latest evaluations have raised concerns about additional Hephaistos candidates.

Some may be affected by active galactic nuclei—bright central regions of distant galaxies powered by matter falling toward supermassive black holes. Others may involve faint background sources, unusual dust environments, or blended photometry. These are all natural phenomena, but they can be spectacularly bright in infrared light.

This is the central challenge of technosignature astronomy: nature is wildly inventive.

The universe has dusty galaxies, forming stars, collapsing black holes, disks of rubble, stellar flares, brown dwarfs, and line-of-sight coincidences. Any one of them can produce something that initially looks artificial if you only have a limited number of measurements.

So, after the newest follow-up work, the honest conclusion is not that a Dyson swarm has been found. It has not.

The honest conclusion is that researchers identified an interesting candidate set, and at least some of its most compelling members now appear to be false positives caused by background galaxies. The remaining objects deserve scrutiny, but no candidate currently meets the standard for a credible extraterrestrial megastructure detection.

That may sound anticlimactic. It shouldn’t. Ruling things out is progress. It makes the next search smarter.

6. Future surveys will look for stronger, harder-to-fake signatures

The next generation of searches will be more sophisticated than simply asking whether a star is “too infrared-bright.”

Researchers are increasingly combining several tests:

• Position on the Hertzsprung–Russell diagram: A star’s luminosity and temperature should fit plausible stellar physics. A system that appears too cool for its luminosity could be worth investigating.
• Detailed infrared spectra: Dust has chemical fingerprints, including silicate features. A smooth, engineered-looking thermal spectrum would be more unusual.
• Variability: Natural disks, galaxies, and active black holes can change in characteristic ways over time. A swarm may produce a different variability pattern.
• High-resolution imaging: This is essential for ruling out unrelated background galaxies.
• Radio and optical observations: A real technosignature claim should survive checks across many parts of the electromagnetic spectrum.

The Project Hephaistos website also places these candidate searches in a broader context: previous work has mainly established upper limits on how common conspicuous Dyson-like systems could be in nearby stellar populations.

In plain English: if civilizations are building massive, warm, easily detectable swarms around many nearby stars, they are not doing it in a way that jumps out of present survey data.

But that still leaves plenty of possibilities. They could be rare. Their waste heat could be cooler or hotter than assumed. Their systems could be incomplete, intermittent, hidden, or built around stars we have not studied well enough. Or, naturally, they might not exist at all.

7. The real lesson is that the search has become scientifically mature

The most exciting thing about these results is not that astronomers have found aliens. They haven’t.

It’s that the hunt for Dyson swarms is becoming a real observational science rather than a thought experiment. Researchers can now scan millions of stars, identify anomalous infrared systems, quantify false-positive rates, and use telescopes like JWST to test the best candidates.

That’s a major shift.

The latest infrared surveys have produced a small number of initially compelling candidates, especially around red dwarfs. Follow-up has already shown that at least two prominent signals came from background galaxies rather than megastructures. Other candidates remain under investigation, but none provides confirmed evidence of extraterrestrial technology.

Still, there’s something quietly thrilling about that. We are learning how to search the Milky Way not merely for planets or chemical traces of life, but for the thermal footprints of civilizations that might have learned to use the energy of a star.

And if one day an infrared anomaly survives every mundane explanation? Well, that will be one very warm mystery.

Suazo, M., Zackrisson, E., Mausam, K., Amiri, A., Korn, A. J., Wright, J. T., & Bondareff, S. (2024). Project Hephaistos – II. Dyson sphere candidates from Gaia DR3, 2MASS, and WISE. Monthly Notices of the Royal Astronomical Society, 531(1), 695–707. https://doi.org/10.1093/mnras/stae1108

Amiri, A. (2026). Searching for Dyson spheres: Modeling their footprint on the H-R diagram and a four-test diagnostic checklist. Universe, 12(2), 142. https://doi.org/10.3390/universe12020142

Bondareff, S., Zackrisson, E., & Suazo, M. (2026). High-resolution follow-up of Project Hephaistos candidates: Resolving WISE contaminants with the James Webb Space Telescope. arXiv preprint arXiv:2607.09460. https://arxiv.org/abs/2607.09460