Astronomers used to assume galaxies formed first and slowly grew the black holes at their centres. But JWST is now finding objects in the early universe where the black hole appears to have arrived first — already enormous before a full galaxy had grown around it.

The old classroom version of galaxy formation had an appealing order to it. First came the galaxy: gas cooled, stars formed, structure assembled. Then the black hole at the centre grew with it.

JWST has made that sequence harder to treat as the default story for the early universe. In several objects seen within the first billion years after the Big Bang, the central black hole appears too large for the surrounding galaxy, at least when judged against the relationship seen in nearby galaxies. The cautious reading is not that galaxies no longer matter. It is that some early black holes may have started large, grown fast, or both.

The nearby universe gave astronomers a tidy rule

In the local universe, large galaxies and their central supermassive black holes tend to scale together. The black hole is massive, but it is usually a small fraction of the stellar mass in the host galaxy. That relationship helped shape a common picture of co-evolution: galaxies grow through gas inflow, star formation and mergers, while their central black holes grow through accretion and their own mergers.

That picture was never meant to be a simple law. It was an empirical pattern, drawn mostly from galaxies observed long after the universe had settled into more recognisable structures. The early universe was always the harder test, because the relevant light had been stretched into infrared wavelengths and the objects themselves were faint.

JWST changed the observational balance. Its infrared sensitivity lets astronomers take spectra of galaxies and compact red sources from the first few hundred million years of the universe’s history, separating starlight, gas emission and active galactic nucleus signatures in ways that were previously out of reach.

UHZ1 put the problem in unusually plain form

One of the clearest examples is UHZ1, a distant galaxy seen through the foreground galaxy cluster Abell 2744. In a November 2023 release, NASA’s Chandra X-ray Center described UHZ1 as a galaxy 13.2 billion light-years away, seen when the universe was only about 3 percent of its current age. The discovery used Chandra X-ray data and JWST infrared observations.

The reason UHZ1 matters is not simply that it is far away. Chandra detected X-rays from material close to a growing supermassive black hole, while Webb measured the distant galaxy in infrared light. According to the Chandra release, the team estimated the black hole’s mass at between 10 million and 100 million solar masses, similar to the estimated mass of all the stars in its host galaxy.

That ratio is the uncomfortable part. Chandra’s release contrasts it with nearby galaxies, where central black holes usually contain only about a tenth of a percent of the stellar mass of their hosts. UHZ1 therefore looks less like a mature galaxy that slowly fed a central black hole, and more like an early system in which the black hole was already a major part of the mass budget.

The Nature Astronomy paper led by Akos Bogdan and the related interpretation by Priyamvada Natarajan and colleagues treat UHZ1 as evidence for a heavy-seed pathway, in which some early black holes formed directly from the collapse of large gas clouds rather than from the remnants of the first stars. That is still a model-dependent interpretation, but it is the kind of interpretation JWST was expected to test.

The little red dots sharpen the question

UHZ1 is not alone in creating trouble for the older sequence. JWST has also uncovered compact, red objects often called little red dots. Their nature is still debated, but many show spectral features that look like active galactic nuclei: gas moving at high speeds around accreting black holes.

In March 2024, Jenny Greene and colleagues published an Astrophysical Journal paper using JWST UNCOVER spectroscopy to examine red sources at redshift greater than 5. The paper, available as an arXiv preprint, argued that active galactic nuclei were surprisingly common in that population. It was an early indication that compact black-hole activity might be much more frequent in the young universe than pre-JWST surveys had suggested.

A later Nature paper led by Ignas Juodžbalis reported a dormant, overmassive black hole in the early universe, based on JADES data. The object lies at redshift 6.68. The authors estimated a black hole mass of about 400 million solar masses and a black-hole-to-stellar-mass ratio of roughly 0.4, about 1,000 times above the local relation.

If the mass estimate and host-galaxy interpretation hold, the black hole was not a minor central feature waiting for the galaxy to catch up. It was already an outsized component in a young system.

The newest case is even more stripped down

The most direct version of the same problem comes from a little red dot behind Abell 2744. In a 2025 paper now linked to a Nature DOI, Juodžbalis and colleagues reported a direct black hole mass measurement in a strongly lensed object at redshift 7.04.

The authors used lensing and deep spectroscopy to measure gas rotation close to the centre of the object. They argued that the rotation is best explained by a point mass of about 50 million solar masses, with little room for a large stellar host. In conservative terms, they inferred that the black hole mass is more than twice the stellar mass.

This is where the phrase “black hole first” becomes tempting. The paper describes the object as a massive black hole seed caught in an early accretion phase. If that reading is right, the black hole is not merely ahead of its galaxy. It is the dominant structure around which the galaxy has not yet fully assembled.

But the caution matters. Little red dots have been difficult to interpret. Some black hole masses are inferred from broad emission lines, and those estimates depend on assumptions about geometry, gas motion and whether lower-redshift calibrations apply in these compact early systems. The direct dynamical case is stronger than many earlier estimates, but the population as a whole is not solved.

Why the first seed matters

The formation problem is simple to state. The universe was young, but the black holes are already large. If the first black holes began as the remnants of massive stars, they would need extremely efficient growth to reach millions or billions of solar masses so early. That may happen in bursts, especially if accretion temporarily exceeds the Eddington limit, but it is demanding.

Heavy-seed models start with a larger initial object. A direct-collapse black hole, for example, could form from a massive gas cloud and begin with tens of thousands or perhaps hundreds of thousands of solar masses. That does not remove the growth problem, but it makes the early timeline less strained.

The JWST evidence does not yet choose one pathway for every early black hole. UHZ1 points toward a heavy seed. GN-z11, analysed in a Nature paper led by Roberto Maiolino, shows an accreting black hole only about 400 million years after the Big Bang, with properties consistent either with heavy seeds or with smaller seeds growing through intense accretion episodes. The little red dots add a broader and stranger population, but their interpretation is still under active argument.

What has changed is the burden of explanation. The early universe is offering compact systems where black-hole growth appears to be at least as advanced as stellar assembly, and sometimes ahead of it.

The sequence may run both ways

The safer conclusion is not that the black hole always came first. It is that the early universe may have had more than one route into the galaxy-black-hole partnership seen later.

In some places, galaxies may have built central black holes gradually. In others, a massive seed may have formed early, pulled gas inward, powered an active nucleus, and then influenced how the surrounding galaxy grew. In still others, bursts of accretion and periods of dormancy may have made a black hole look too large before later star formation narrowed the gap.

JWST’s contribution is not a neat replacement story. It is a set of objects that refuse to fit the older order cleanly. The next test will be whether deeper spectra, X-ray observations, lensing measurements and future gravitational-wave detections can separate genuinely early heavy seeds from black holes that only look overgrown because of how young, compact and dust-filled their hosts are.

For now, the old sequence has become a question rather than a rule. In the first galaxies, the centre may not always have waited for the galaxy to arrive.