Picture a star giving off more heat than it should. Not a flare, not a flicker, just a steady extra glow in the infrared, the part of the light spectrum our eyes cannot see but the right telescope can. For a small, dim star, there is no easy natural reason for that warmth. Unless something enormous is wrapped around it, quietly catching starlight and breathing the leftover heat back into space.
For sixty years that idea, a civilization building a shell of solar collectors around its star, had almost no way to be tested. Now that is changing fast. A team just ran its best candidates through the most powerful telescope ever flown, and a wave of new machines is lining up behind it.
The first headline of that new era is not a discovery. It is a disqualification. And it is exactly what a science growing up is supposed to look like.
20 BILLION
cosmic objects the Vera Rubin Observatory will scan for the flicker of a passing megastructure.
Here is what happened
It came in two steps, two years apart.
The scan, 2024. A search called Project Hephaistos, led by astronomer Erik Zackrisson at Uppsala University, combed roughly five million Milky Way stars using three sky surveys: Gaia for positions and distances, plus 2MASS and WISE for infrared heat. Seven stars passed every filter, all of them red dwarfs, the small, cool, dim stars that make up most of the galaxy.
The test, July 2026. The team aimed the James Webb Space Telescope, the sharpest infrared observatory ever launched, at the two strongest suspects, Candidate D and Candidate E.
The verdict. Neither was wrapped in alien engineering. In each case the extra glow came from a separate galaxy sitting almost exactly behind the star, so precisely lined up that WISE's blurrier eye had smeared the two into a single dot.
What was hiding. Behind Candidate D, a rare, dust-choked galaxy called a Hot DOG, lit by a feeding supermassive black hole. Behind Candidate E, a galaxy forming stars fast enough to heat its own dust into a bright infrared shine.
The tally. Two of the seven are now explained. Five remain open, and scientists expect ordinary astrophysics will account for those too.
How you catch a civilization by its waste heat
The trick rests on a law no engineer can beat.
Waste heat is unavoidable. In 1960, physicist Freeman Dyson noted that any civilization capturing a large share of its star's energy runs into thermodynamics. No machine is perfectly efficient, so whatever the collectors do not use has to leave as heat, the same warmth a laptop or a lightbulb gives off, just on a stellar scale.
It glows in the infrared. For a structure a few hundred degrees above absolute zero, that heat shows up mainly as mid-infrared light. A star with more mid-infrared than its size and temperature should allow has an "infrared excess," and that is exactly what Project Hephaistos hunts for.
Why red dwarfs. A warm signal is easier to spot against a faint, cool star, and a swarm around a small star needs far less material to build.
How JWST settled it. It did what the older WISE survey could not: it separated the light. At longer wavelengths the star faded while a second point beside it brightened, the mark of two objects, not one, and a follow-up spectrum confirmed each companion was a galaxy. The method worked. The candidates simply were not aliens.
The machines that turn the hunt industrial
Here is what makes this a 2026 story rather than a 1960 one. For most of its history the search was starved: a handful of candidates, examined one at a time, on instruments never built for the job. Now the tools are arriving together, and a rulebook is arriving to point them.
Vera Rubin adds time. The observatory in Chile began its ten-year survey in June 2026, filming almost the entire southern sky every few nights. A partial Dyson swarm drifting in front of its star would cause strange, irregular dimming, the signature that made the famous "Tabby's Star" a sensation a decade ago. Rubin turns that lucky one-off into a systematic sweep of about twenty billion objects, pouring out roughly twenty terabytes a night, with machine-learning anomaly detectors taking the first pass.
Roman adds reach. NASA's Nancy Grace Roman Space Telescope, launching on August 30, 2026, delivers Hubble-sharp images across a patch of sky about a hundred times wider, in infrared, so it can hunt the same waste-heat excess across vastly more stars. Its separate microlensing survey gives a second, independent probe.
A new lensing trick. When a massive object passes in front of a star, its gravity magnifies the starlight in a precise pattern, and a solid or semi-solid megastructure would bend that pattern in a telltale way. Published in late 2026, the idea could catch structures too cold to glow, a clean complement to the waste-heat method.
A four-test checklist. Proposed in 2026 and timed to this moment, a genuine detection would need all four: an impossible spot on the chart of a star's brightness versus temperature, an infrared glow with none of the chemical fingerprints of natural dust, irregular dimming unlike any known variable star, and a light signature matching nothing in the catalog of stars, dust, or galaxies. Nothing found so far clears all four.
The 2040s leap. NASA's next flagship, the Habitable Worlds Observatory, is a six-to-eight-meter telescope built to directly image at least twenty-five Earth-sized planets around nearby stars and read the chemistry of their air. It is designed to detect biosignatures, the signs of life, and technosignatures, the signs of technology, such as an industrial pollution pair of carbon dioxide plus nitrogen dioxide.
More angles still. Europe's proposed LIFE mission would chase the same atmospheric fingerprints with a formation of spacecraft flying as one giant infrared instrument, while Breakthrough Listen and the coming Square Kilometre Array keep listening for deliberate radio signals, the other way a civilization might give itself away.
The through-line is a shift in kind, not degree: from a few candidates checked by hand to billions of stars watched nonstop, sifted by AI, and eventually cross-examined by space telescopes built to read the air of other worlds.
Why it matters
The strange part is that even finding nothing would count as a result.
Rigor you can see. A team disqualifying its own best leads, in public, on the best telescope ever flown, is half the story.
Numbers beat anecdotes. Survey enough stars and a non-detection becomes real evidence: hard limits on how rare, or absent, energy-hungry civilizations are. That is a direct grip on the Fermi question, the puzzle of why a galaxy this old and vast looks so quiet.
Scale is the whole point. Five million stars sounded like a lot, but the Milky Way holds well over a hundred billion. Rubin, Roman, and one day the Habitable Worlds Observatory widen the net by orders of magnitude, and turn a search that ran on anecdotes into one that can run on statistics.
The honest catch
Even a clean result and a bright roadmap come with limits worth naming plainly.
It is two of seven, and a preprint. Only the two strongest candidates have been examined this closely, five remain open, and the findings still await full peer review. Catching two fakes proves the filter can spot impostors, not that it would recognize a real signal, because no confirmed example exists anywhere to test it against.
The toolkit arrives staggered. Rubin and Roman are here or imminent, but the Habitable Worlds Observatory is a 2040s concept not yet built or fully funded, and LIFE is still a proposal. Each new instrument also brings a fresh flood of natural false alarms to rule out first.
It only tests certain ideas. Waste heat, odd dimming, atmospheric pollution. A civilization that manages its heat cleverly, or signals in a way nobody has thought to look for, would stay invisible to every machine on this list.
EDITOR'S TAKE
The exciting part is not this month's non-detection, it is that the hunt is leaving its data-starved era for good. For sixty years astronomers had one Tabby's Star and a scatter of oddities. Within a decade they will have billions of stars watched night after night, sifted by AI, and later a telescope built to read the atmospheres of other Earths. Watch Rubin's first anomaly catalogs and Roman's wide infrared survey now, and watch the harder fight to actually fund and build the Habitable Worlds Observatory for the 2040s. The lesson only grows sharper as the machines improve: the better the instrument, the higher the bar for proof. The day a candidate clears every test on the best telescope ever flown and still cannot be explained away, that is the story. This is the field quietly building the tools to make that day possible.
Quick questions
What exactly is a Dyson sphere?
A Dyson sphere is a hypothetical structure, or more realistically a swarm of many separate collectors, that an advanced civilization might build to capture most of a star's energy. Physicist Freeman Dyson proposed the idea in 1960, though he treated it partly as a thought experiment. Modern versions imagine a loose swarm rather than a solid shell, since a solid shell would not be stable around a star. The giveaway it would leave is a warm infrared glow, the waste heat left over once the civilization uses the starlight it captures.
How will future telescopes actually spot one?
Three different ways, which is why the coming decade matters so much. The Vera C. Rubin Observatory watches how billions of stars flicker over time, so it can catch the strange, irregular dimming a passing swarm of collectors would cause. NASA's Roman Space Telescope scans huge fields in infrared to hunt waste-heat glows, and its microlensing survey can flag the odd way a solid structure would bend a star's magnified light. Further out, the Habitable Worlds Observatory aims to directly image nearby Earth-sized planets and read their atmospheres for industrial pollution, a chemical sign of technology. A four-test checklist, published in 2026, sets the standard all of them would have to clear.
Did the James Webb Space Telescope find aliens?
No, and the scientists were careful to say so themselves. JWST examined the two most promising Dyson-sphere candidates out of an original list of seven and found both were false alarms caused by distant galaxies sitting almost exactly behind the stars as seen from Earth. One hidden galaxy is fed by a supermassive black hole and the other is forming stars unusually fast, and both naturally glow in infrared. Five candidates from the original search have not yet been examined at this level of detail, but nothing in this study points to alien technology.
Sources
Project Hephaistos IV: JWST Observations of Two Dyson Sphere Candidates: Primary paper: JWST rules out Candidates D and E as background galaxies, not megastructures.
Project Hephaistos II: Dyson Sphere Candidates from Gaia DR3, 2MASS, and WISE: The original 2024 study that found seven red dwarf candidates among five million stars.
Search for Artificial Stellar Sources of Infrared Radiation, Freeman J. Dyson, Science, 1960: The founding paper proposing a search for alien waste heat in infrared light.
Dyson Sphere Hunt Gains Four-Test Checklist as Webb, Rubin, Roman Converge: The 2026 detection checklist and the three-telescope convergence.
Microlensing Signatures of Dyson Sphere-like Structures as Technosignatures: New 2026 proposal to catch megastructures through their gravitational lensing signature.
NASA lays groundwork for the Habitable Worlds Observatory in the 2040s: The future flagship built to image other Earths and read their atmospheres for bio and technosignatures.
Projections of Earth's Technosphere: Observing Technosignatures on Terrestrial Exoplanets: Peer-reviewed basis for the industrial-pollution technosignature that HWO could look for.
Frontier Signal explains frontier technology in plain English. Company-reported and preprint figures should be independently verified. This is general information, not investment advice.
