3I/ATLAS: the third visitor, and the alien question

Scene · reconstructed from the paperworkPicture it. A survey telescope in northern Chile, doing the dull and important job of watching for rocks that might hit us. On 1 July 2025[nasa-facts] its observations go to the Minor Planet Center, and something on the list is moving wrong.

The telescope is ATLAS, the Asteroid Terrestrial-impact Last Alert System, in Río Hurtado[esa-faq]. It is paid for by NASA and it exists for planetary defense. What it catches is a speck about 670 million kilometres from the Sun, inside Jupiter's orbit[nasa-facts].

The speck is too fast. At discovery it is doing about 221,000 km/h[nasa-facts], and its path does not close into an orbit. It is on what astronomers call a hyperbolic trajectory: moving too quickly for the Sun to hold, passing through once. Observers around the world confirm that it came from outside[esa-faq].

That makes it the third such object ever seen. ‘Oumuamua came first in 2017, then 2I/Borisov in 2019[esa-faq]. So it gets the name 3I/ATLAS: third, interstellar, found by ATLAS. Its other name, C/2025 N1 (ATLAS)[esa-faq], says something the first visitor's name could not. The C stands for comet. Unlike ‘Oumuamua, this one shows activity quickly. Hubble catches a dust plume from its sunward side and the hint of a tail[esa-faq], behaviour familiar from comets at home.

How big? Using Hubble, astronomers bracket the nucleus at no less than 440 metres and no more than 5.6 kilometres[nasa-facts]. That is a wide range.

Is it dangerous? No, and this is settled early. NASA's planetary defense people say right away[defensescoop-nov] that it poses no threat. It will come no closer than about 270 million kilometres[nasa-facts], nearly twice the distance from Earth to the Sun, and at closest approach to the Sun it will be on the far side from us[nasa-facts].

Then there is the part that makes astronomers sit up. Its speed suggests age. ESA says it has probably been drifting through space for billions of years[esa-faq]. A telescope built to protect us has found something that lets us look at another star's leftovers without leaving home.

The window is short. By the time anyone has a plan, it is already falling toward the Sun at speed. ESA is frank that it is far too quick to send a spacecraft to visit[esa-faq]. So the world's observatories do the next best thing: they all turn and look.

What follows is the most closely watched comet passage of its kind, and the paper trail runs through nearly every space agency that owns a telescope.

Webb looks in August 2025 and finds the coma releasing carbon dioxide, water, carbon monoxide, carbonyl sulphide and water ice[esa-faq]. Those are comet ingredients. Hubble, SPHEREx and others join in, and NASA's list grows to include Hubble, Webb, TESS, Swift, SPHEREx, the Perseverance rover and more[defensescoop-nov].

On 3 October 2025[esa-faq] the comet passes Mars at about 29 million kilometres. ESA's Mars Express and ExoMars Trace Gas Orbiter watch, and the orbiter's data improves the predicted path by a factor of ten[esa-faq]. A spacecraft at Mars has helped navigate a visitor from another star. It sounds grander than it is, and it is still rather grand.

Near the end of October it reaches perihelion, its closest point to the Sun, about 1.4 astronomical units out, just beyond Mars's orbit[nasa-facts]. NASA gives 30 October and ESA gives 29 October[esa-faq], so the date depends on whose page you read. At top speed it reaches about 246,000 km/h[nasa-facts].

Here the story gets a human complication. The comet goes behind the Sun from our point of view, and at the same moment the US government shuts down. From 1 October to mid-November[defensescoop-nov] NASA cannot release its images. The shutdown is record-breaking[defensescoop-nov], and speculation fills the gap.

On 19 November[defensescoop-nov] NASA's senior officials finally hold the briefing and release the pictures. Associate administrator Amit Kshatriya says he was struck by how excited people were. Nicky Fox says what everyone came to hear: it behaves like a comet, and they have seen no techno-signatures[defensescoop-nov]. Then she adds the better part, that the differences are what make it interesting, since it may predate our Sun.

A quieter line in the same report is worth a pencil note. The Pentagon's UFO office, AARO, is following the object, but a defense official says it falls outside AARO's mission, as it is assessed to be a comet and so is not UAP[defensescoop-nov]. The people paid to look into unidentified objects decline it, because it is identified.

Later that month Hubble comes back and takes the picture in this file, on 30 November[nasa-hubble-dec], and X-ray telescopes make it the first interstellar comet seen in X-ray light[esa-faq]. On 19 December[esa-faq] it makes its closest pass of Earth, and nothing happens, which was the forecast. After that it heads out toward Jupiter's orbit in spring 2026[defensescoop-nov].

Then the slow part begins, the part that never makes the news: the data goes to laboratories and papers. Webb is pointed at it again in December 2025[esawebb-june], as it leaves. That session will produce the strangest results of all, and none of them involve engines.

Now the argument. It starts within two weeks of the discovery.

On 16 July 2025[hibberd-2025] three authors post a preprint with a title built for the internet: Is the Interstellar Object 3I/ATLAS Alien Technology? One of them is Abraham Loeb, the Harvard astrophysicist known as Avi, who also co-wrote the light-sail paper on ‘Oumuamua[bialy-loeb-2018]. He is a well-known scientist, and his papers get read.

Read the abstract closely, because most retellings don't. Its own first move is to call itself largely a pedagogical exercise[hibberd-2025]. It then hypothesises that the object could be technological, and possibly hostile, as the “Dark Forest” answer to the Fermi paradox would predict[hibberd-2025]. It notes that the path passes surprisingly close to Venus, Mars and Jupiter. It suggests that being hidden behind the Sun at perihelion would allow a clandestine braking manoeuvre[hibberd-2025], and that an optimal intercept of Earth would arrive in late November or early December 2025.

That last line is the useful one. It is a test.

The comet does pass behind the Sun, as the paper noted. NASA goes quiet in the shutdown. For some weeks the idea has room to run. Then the images arrive, and December passes with no intercept. NASA says it has seen no techno-signatures[defensescoop-nov]. The object reaches its closest point to Earth on 19 December and carries on.

The same day Loeb goes on NewsNation, and it is worth hearing what he actually says, since people on both sides shorten it. He says 3I/ATLAS did not maneuver or display unusual activity near Earth[loeb-newsnation]. Asked directly, he says that given all the data, he agrees it is most likely natural[loeb-newsnation], but that much is still not understood. He holds on to one thing in particular: a jet he describes as pointing at the Sun and stretching about a million kilometres, which he says was never seen in Solar System comets. He also sets himself a further test. If nothing unusual happens as it nears Jupiter on 16 March 2026[loeb-newsnation], he says, he would conclude it must be natural.

We do not mock this, and we do not think you should. Asking the question was legitimate. A scientist named a falsifiable prediction, the prediction failed, and he said so on television. That is how this is meant to work, even if the journey there was loud. The critics have a point as well: a “possibly hostile”[hibberd-2025] headline travels further than a later, quieter statement, and it was attached to a paper that called itself an exercise.

Notice also what the comet did not need. Nothing about its behaviour required a new category. What it required was better chemistry, and that was already on its way.

The real surprise came in 2026, from the isotopes. It turned out that the wildest fact about this object is true, and has nothing to do with aliens. It may be older than the Sun.

What is still open is much more interesting than the question that made the headlines.

In May 2026 a University of Michigan team reports on the comet's water, measured with ALMA. The share of deuterium, or heavy hydrogen, is about 30 times that of Solar System comets and 40 times that of Earth's oceans[salazar-2026]. It is the first time this kind of analysis has been done on an interstellar object.

In June a second team confirms the picture with Webb, in Nature. They measure water D/H of 0.98 percent, and carbon isotope ratios unlike anything in the Solar System or nearby interstellar clouds[cordiner-2026]. Interpreted through models of how the Galaxy's chemistry changes over time, the carbon implies the comet may have formed as long as 12 billion years ago[cordiner-2026]. The press release puts it at 10 to 12 billion years[esawebb-june], when star formation was at its height. Our Sun is about four and a half billion years old.

Hold that thought. This object was already ancient before there was an Earth to pass.

In October a third result arrives, in Nature Astronomy. Webb picks out deuterated methane, at 3.18 ± 0.30 percent, higher than any methane value in the Solar System and 13 ± 2 times that of comet 67P[roth-2026]. The result points to ice formed in intense cold, around 10 degrees above absolute zero[physorg-oct]. But the careful reading matters. It offers a plausible history, not an identified birthplace[physorg-oct], and the few chemical models that report such values fall short of what was measured. So the comet has beaten the models. That is a real open problem, and a good one.

A warning about press numbers: some reports of that paper give 14 times and about 3.33 percent[scinews-oct]. The paper's own abstract says what we quote here.

There are other loose threads. The estimated age leans on models and on where in the Galaxy the comet formed[cordiner-2026], and Juice's data only reached Earth in February 2026 and is still being analysed[esa-faq]. The jet Loeb points to is his reading, and we have not found a peer-reviewed paper that treats it as an anomaly needing new physics. We have also not found a clear account of what he concluded after 16 March. Both sit under To Verify.

And there is the larger question. Three visitors in eight years is a very small sample. NASA expects more to be found as better instruments arrive[defensescoop-nov]. One of the researchers' reasons for caring is stated plainly by Stefanie Milam: this is about how common, or uncommon, the conditions for life are[esawebb-june]. Martin Cordiner says the same from the other end. It tells us how unusual our own Solar System may be[esawebb-june].

3I/ATLAS has now left at the speed it came in[nasa-facts]. It is not coming back. What it left is in the file below: a set of ratios, measured from a long way away, from somewhere we cannot yet name.