JWST found an object that should not exist — and rewrote what we know about the early universe

2026-08-19 · by Mira Okafor · Newsletter

There is an object in the constellation Cetus, 660 million years after the Big Bang, that looks like an enormous star. It glows red. It is roughly the size of the solar system. And according to a paper published in Nature on August 12, 2026, it is being powered by something no star should have at its center: a black hole roughly 100,000 times the mass of the Sun, wrapped inside a cocoon of hydrogen and helium so dense it mimics a star from the outside. The lead author, Rohan Naidu of MIT's Kavli Institute for Astrophysics and Space Research, calls it a "black hole star." The paper, with co-authors including MKI Director Robert Simcoe and MIT researcher Wendy Sun, describes the first confirmed observation of an object that existing astrophysical models had not predicted — one that emerged not from a targeted search for new physics, but from a routine survey of the early universe looking for something else entirely.

How JWST found an object without a name

Naidu and his colleagues were not hunting for black hole stars. They were running a survey they named "Mirage or Miracle" (MoM), using JWST to identify the earliest and most distant galaxies — ones that formed in the first few hundred million years after the Big Bang. The method is straightforward in principle: look very far away, which means very far back in time, and count the galaxies. In practice, it has produced one of the most productive surprises in modern astrophysics.

"We started looking for extremely bright galaxies in the early universe," Naidu explained in an MIT News interview. "What we found was that what looks like an extremely bright early galaxy — a 'miracle' — in some cases actually could be a 'mirage.'" The mirage in this case was MoM-BH*1: an object in the JWST image archive that was both extraordinarily red and extraordinarily bright, and that did not match any known class of galaxy, quasar, or star.

The team ran the object through spectral analysis — splitting its light into component wavelengths — and found something unexpected. The spectrum showed what astronomers call a Balmer break, a sharp drop in light below a certain wavelength that typically indicates dense gas in the atmosphere of an old star. "The break we observed in this object is the deepest break we have ever observed in any object," Naidu said, "ruling out 'ordinary' stars as the source." And yet the object was far too luminous to be an old star. Something else was generating that energy.

The physics of something that should not work

The standard energy sources in astrophysics are two: nuclear fusion in a star's core, or accretion onto a black hole. Fusion produces stars. Accretion onto a black hole produces quasars and active galactic nuclei — enormously bright structures, but ones that sit at the centers of galaxies, not scattered across intergalactic space as lone objects.

MoM-BH*1 appears to be something between those two categories. The leading model from Naidu's team is a central black hole — about 100,000 solar masses — surrounded by an extremely dense envelope of hydrogen gas. That envelope is so optically thick that light cannot pass through it cleanly. From the outside, the structure looks like a single enormous star. From the inside, it is a black hole generating energy through accretion, surrounded by material that mimics stellar fusion signatures.

The object also has an unusual chemical composition. Its spectrum shows almost no metals — no elements heavier than hydrogen and helium. This is significant. Early-universe gas is metal-poor by definition, but most early-universe objects still carry some metal content from previous stellar generations. MoM-BH*1 is essentially pure. "It was truly singular in so many ways," Naidu said.

Computer simulations run by the team confirmed the basic picture: a dense hydrogen cocoon surrounding an accreting black hole produces both the extreme redness (the cocoon absorbs and reddens light from the inner region) and the deep Balmer break (the dense gas at the surface absorbs the shorter wavelengths). Simcoe, who worked on the spectroscopic analysis, described the process: "We asked: could you make something that red using just hydrogen, without any dust? To our surprise, it turns out you can, if you have an extremely dense screen of hydrogen, so dense that it looks more like the surface of an enormous star than a wispy interstellar nebula."

The energy output is the most striking confirmation. MoM-BH*1 produces roughly 100 billion times more energy than any known individual star can generate through fusion. That output level is consistent with black hole accretion — but the object has none of the other signatures typically associated with active galactic nuclei, because the accretion region is buried deep inside the hydrogen cocoon.

Why this matters for the whole universe

The discovery has implications beyond the object itself. When the team examined other JWST deep-field images, they found that a significant fraction of the mysterious "little red dots" — small, extremely red objects that appear in nearly every JWST image of the early universe — share spectral characteristics consistent with the black hole star model. "These little red dots seem to be everywhere in the early universe but essentially disappear by the present day," Naidu said. "What exactly these objects are has been one of the most debated topics of the JWST era."

If most little red dots are black hole stars, it changes the story of how the early universe evolved. The leading model for supermassive black hole formation faces a timing problem: to reach the masses we observe at the centers of present-day galaxies, black holes in the early universe had to grow very fast from very large seeds. The standard mechanism — small seed black holes from the first stellar generation, growing by gradual accretion — may not have had enough time. Black hole stars offer an alternative: direct collapse of massive gas clouds into large seed black holes, with the black hole and its cocoon visible as a distinct object before the cocoon disperses and the black hole settles into a galactic center.

"For decades we have anticipated something spectacular must be afoot in the very early universe," Naidu said. "Black hole stars may be the 'something spectacular.' They may be the nascent, swaddled phase that marks the beginning of almost every supermassive black hole's journey."

That is a strong claim, and it will require more observations to confirm. But it is also a claim with a direct path to confirmation: if the little red dots are black hole stars, their numbers and distribution in the early universe can be measured against the prediction. JWST is the instrument that makes that measurement possible.

What JWST keeps finding at the edge of what we understood

The black hole star is not the only recent JWST finding in the same category — a discovery that required the telescope to exist before the discovery could be made. On August 11, an international team using JWST published separate observations of the environment around Sagittarius A, the supermassive black hole at the center of our own Milky Way. The telescope's Mid-Infrared Instrument detected dust and water in the immediate vicinity of IRS 3, an evolved star less than a light-year from Sagittarius A, in conditions astronomers had expected to be too harsh for dust survival. The discovery shows that even in one of the most extreme radiation environments in the galaxy, stellar evolution can continue enriching the local environment — a result with implications for understanding how planetary systems form in regions of high stellar density.

Neither observation would have been possible before JWST. The black hole star required JWST's sensitivity to detect a faint, red, distant object and measure its spectrum with enough precision to rule out known alternatives. The Sagittarius A* observations required JWST's mid-infrared capability to detect dust signatures where ground-based telescopes are blinded by atmospheric interference. In both cases, the telescope did exactly what it was designed to do: see things that were previously too faint, too distant, or too obscured to detect. In both cases, what it found was not what the surveys were looking for.

That is how science works sometimes. JWST was designed primarily to find the first galaxies. In the process of doing that, it is finding objects we did not have categories for before. The black hole star is not an answer to the question the MoM survey asked. It is a much more interesting result: evidence of a category of astrophysical phenomenon that our models of 2022 did not predict, discovered because we finally built something precise enough to see clearly into the early universe.

The Nature paper (Naidu et al., 2026, doi: s41586-026-10846-4) carries the cautious language of peer-reviewed science — the authors describe their interpretation as the leading model consistent with the data, not an established fact. That caution is warranted. But the object is real, the spectrum is real, and the problem of what is powering it has a small number of candidate explanations. One of those candidates is a new kind of astrophysical object. JWST has a habit of producing those.

---

Sources

Mira Okafor
Contributor