Rogue Black Hole Found on the Outskirts of a Distant Galaxy

Astronomers at the University of Maryland say they have caught a supermassive black hole that does not sit at the center of its galaxy, but 9.3 kiloparsecs away from it, roughly 30,000 light years. It gave itself away with a single short flare, when it tore apart a star that strayed too close. The paper, titled “TDE 2025abcr: A Tidal Disruption Event in the Outskirts of a Massive Galaxy,” was published on July 27, 2026, in Astrophysical Journal Letters, volume 1006, article L57. It is the first time a dormant black hole has been detected this far from the nucleus of its host galaxy.

Wandering black holes are nothing new in theory. Models of galaxy collisions and mergers have long predicted that some black holes end up ejected to the outskirts, far from the dense center. The trouble was that such an object typically stays invisible. When a black hole is not feeding on matter, it emits no light, and for telescopes on Earth it simply does not exist.

How the machine learning program found the flare

The data comes from the Zwicky Transient Facility, a sky survey that uses two telescopes at the Palomar Observatory in San Diego County, California, to scan the entire northern sky every two days. That survey logs hundreds of thousands of variable events across the sky every night, which makes sifting through them by hand unworkable.

So the team built a machine learning program trained to recognize the distinctive light curve produced when a black hole rips a star apart, an event known as tidal disruption. Such events are well documented at the centers of galaxies, but this program was told to look for the same pattern anywhere in the sky. It went live in August 2025, and by November 2025 it had already flagged the candidate that led to the discovery.

“We weren’t entirely sure we’d manage this fast, so it’s incredible that we found it this quickly.”

That was Robert David Stein, a Neil Gehrels Postdoctoral Fellow at the Joint Space-Science Institute and the paper’s lead author. The candidate was then followed up from the ground with the Lowell Discovery Telescope and the infrared instrument RIMAS, which had come online in June 2025.

An object with no galaxy around it

The black hole sits in the galaxy WISEA J014656.04-152214.7, about 750 million light years away, in the constellation Cetus. Its mass is comparable to the black hole at the center of the Milky Way: measurements of Sagittarius A* put that one at roughly four million solar masses. The flare produced by the star’s destruction was briefly brighter than ten billion Suns.

The 9.3 kiloparsec distance is worth pausing on. Our Sun sits about 8.2 kiloparsecs from the center of the Milky Way, so slightly closer in. This black hole, then, sits at roughly the same distance from its galaxy’s center that the Solar System sits from ours, except that around it there is nothing resembling the surroundings an object of that mass ought to have.

“It surprises me that such a massive black hole is sitting outside a galaxy. There should be a Milky Way like object around it, and that is quite clearly not the case.”

That is how Sylvain Veilleux, a University of Maryland astronomy professor and study co-author, described the situation. The researchers suspect the position is the result of an earlier galaxy collision. One possibility is that a larger galaxy swallowed a smaller companion and kept stripping stars from it until only the nucleus was left. Another is that two black holes were already orbiting at the center of one galaxy, and the arrival of a third, from a fresh merger, kicked the smallest one out. Continued observation of the same event should show which explanation holds up.

How common are wandering black holes?

That is the question the team is now asking, and Stein put it this way:

“The key scientific question we want to answer is how common wandering black holes are.”

The answer depends on how many more turn up, and that number should rise sharply once the Vera C. Rubin Observatory in Chile enters full operation, with its ten year sky survey and the largest digital camera in the world.

Suvi Gezari, associate professor of astronomy at the University of Maryland and a study co-author, says:

“This discovery will have a huge impact. It means we’ll find many more examples of wandering black holes, and that we can understand how galaxies and their black holes merge.”

She also offered an assessment of the method itself: “It’s an example of how machine learning is opening up an entirely new field of research.” S. Bradley Cenko and Jillian Rastinejad also worked on the team.

What this finding does not prove

This is a single event. TDE 2025abcr shows that an object of this mass can sit far from a galaxy’s center, but it says nothing about how many such objects exist across the cosmos, since a frequency cannot be estimated from one case. The black hole’s origin has not been established, and the two proposed scenarios remain scenarios until further observations can tell them apart.

The wording also matters here. No galaxy was seen around the black hole where one would be expected, which is not the same as proving there is none. That is a limit of instrument sensitivity, not a measured void. Finally, one hit from the machine learning program says nothing about its miss rate, a figure that will only emerge once enough years of observation have passed through the same filter.