BREAKING NEWS: the interstellar object 3I/ATLAS is anomalously massive and large!
In a new paper (accessible here) that I co-authored with Richard Cloete and Peter Veres, we used data on the motion of 3I/ATLAS — as compiled by the Minor Planet Center between May 15 and September 23, 2025, to set an upper limit on the deviation of 3I/ATLAS from a trajectory sculpted by gravity alone.
The upper limit on residuals is 0.028 arcseconds over a period of several months, implying that the non-gravitational acceleration of 3I/ATLAS was smaller than 15 meters per day squared. The total mass loss rate and outflow speed from its surface were inferred from the Webb Space Telescope data on August 6, 2025 (accessible here). Given that the rocket effect is not noticeable, the mass of 3I/ATLAS must be bigger than 33 billion tons. Consequently, the diameter of its solid-density nucleus must be larger than 5 kilometers. This suggests that 3I/ATLAS is more massive than the other two interstellar objects, 1I/`Oumuamua and 2I/Borisov by 3–5 orders of magnitude, constituting a major anomaly. Given the limited reservoir of heavy elements, we should have discovered of order of a hundred thousand interstellar objects on the 0.1-kilometer scale of 1I/`Oumuamua before finding 3I/ATLAS, yet we only detected two interstellar objects previously.
Our analysis compared the evolution in the observed sky positions of 3I/ATLAS (Right Ascension [RA] and Declination [Dec]) to the trajectory expected from gravity alone. The optical astrometric data span 4,022 individual measurements from 227 observatories worldwide.
The comparison to the expected sky positions provided a net upper limit of 0.028 arcseconds on sky position shifts during the 4.5 months between May 15 and September 23, 2025. A systematic trend is unlikely to be affected by a shift in the centroid of light owing to the development of a bright spot away from the nucleus of 3I/ATLAS. The highest-resolution image of 3I/ATLAS, taken by the Hubble Space Telescope on July 21, 2025 (accessible here), showed an optically-thin anti-tail in the direction of the Sun which persisted during July and most of August (and analyzed here). On August 27, 2025, imaging by the Gemini South telescope revealed the growth of a faint tail from 3I/ATLAS away from the Sun (as reported here). Throughout the entire period, the brightest point remained centered on the nucleus.
Over a short time interval, T, a non-gravitational acceleration, A, results in a positional shift that grows as 0.5*A*T². Given that the distance of 3I/ATLAS from Earth in mid-August was 2.6 times the Earth-Sun separation (as discussed by the SPHEREx team here), our upper limit on the sky position shift translates to an upper limit on the non-gravitational acceleration, A<15 meters per day squared.
Momentum conservation implies that 3I/ATLAS must be massive in order not to be pushed back by the flow of material from its warmed Sun-facing side. The product of the mass loss rate and the outflow speed divided by the upper limit on the non-gravitational acceleration, sets a lower limit on the object’s mass. The Webb telescope data taken on August 6, 2025 (accessible here) implies a total mass loss rate of 150 kilograms per second and an outflow speed of 440 meters per second. Given these values, our upper limit on the non-gravitational acceleration yields a minimum mass of 33 billion tons for 3I/ATLAS. At solid density, this implies a minimum diameter of 5 kilometers for the nucleus of 3I/ATLAS. This minimum diameter is close to the upper bound of the range of 0.44 to 5.6 kilometers inferred from the Hubble telescope data (as reported here). It is an order of magnitude larger than the inferred diameter of the interstellar comet 2I/Borisov (as discussed here).
A large nucleus mass for 3I/ATLAS exacerbates the tension between the interstellar mass reservoir of rocky material ejected from exoplanetary systems and the implied detection rate of interstellar objects of its size, as I derived in a paper published shortly after the discovery of 3I/ATLAS (accessible here).
The passage of 3I/ATLAS near Mars on October 3, 2025 will allow the HiRISE camera onboard the Mars Reconnaissance Orbiter to achieve a pixel resolution of 30 kilometers. The amount of sunlight reflected from the brightest pixel in the HiRISE image will constrain further the nucleus surface area for an assumed albedo value. Additional constraints can be obtained on March 16, 2026, when 3I/ATLAS will pass in proximity to the Juno spacecraft near Jupiter (as proposed in the paper I published with Adam Hibberd and Adam Crowl here and endorsed in a letter by Rep. Anna Paulina Luna here).
The mass of 3I/ATLAS scales with its diameter cubed. If the nucleus diameter of 3I/ATLAS will be found to be larger than 5 kilometers in the HiRISE image, then an origin associated with the interstellar mass reservoir of rocky material will be untenable. An alternative technological origin could explain the rare alignment of the trajectory of 3I/ATLAS with the ecliptic plane (having a random chance of 1 in 500, as discussed here), and the detection of nickel without iron — as found in industrially-manufactured alloys (as reported by the VLT team here).
Given this tight upper limit on non-gravitational acceleration, a future detection of a major maneuver of 3I/ATLAS would suggest propulsion by a technologically manufactured engine.
Is 3I/ATLAS an unusually massive comet with an unusual chemical composition on an unusually rare trajectory or alien technology? In both cases, the object could shed CO2 and H2O ices from material that collected on its frozen surface as it plowed through interplanetary and interstellar space. We should not decide about the nature of 3I/ATLAS based on the chemical composition of its skin, for the same reason that we should not judge a book by its cover.
Hopefully, we will know more in the coming weeks. Stay curious. As Galileo Galilei instructed us, scientific truth is revealed by data, not by authority.
ABOUT THE AUTHOR
Avi Loeb is the head of the Galileo Project, founding director of Harvard University’s — Black Hole Initiative, director of the Institute for Theory and Computation at the Harvard-Smithsonian Center for Astrophysics, and the former chair of the astronomy department at Harvard University (2011–2020). He is a former member of the President’s Council of Advisors on Science and Technology and a former chair of the Board on Physics and Astronomy of the National Academies. He is the bestselling author of “Extraterrestrial: The First Sign of Intelligent Life Beyond Earth” and a co-author of the textbook “Life in the Cosmos”, both published in 2021. The paperback edition of his new book, titled “Interstellar”, was published in August 2024.