A supermassive black hole called RBH-1, weighing at least ten million Suns, appears to have been flung from its galaxy and is now racing through space at nearly 1,000 kilometres per second, leaving a 200,000-light-year wake of newborn stars

A thin streak of gas and young stars trailing a distant galaxy is the wake of a supermassive black hole moving at close to 1,000 kilometres per second, according to a paper by Pieter van Dokkum of Yale University and colleagues, published in The Astrophysical Journal Letters and available as a preprint on arXiv. Using the James Webb Space Telescope, the team found the kinematic signature that had been missing when the object was first reported: a sudden, sharp change in gas velocity at the leading tip, of the kind only a fast, massive body ploughing through thin gas can produce.

That feature runs about 62 kiloparsecs, or roughly 200,000 light-years, and sits beside a galaxy at a redshift of 0.96, meaning its light set out around 7.5 billion years ago. The object has been named RBH-1, and the authors describe it as the first confirmed runaway supermassive black hole. Astronomers have predicted for about half a century that black holes can be flung out of galaxies. Candidates have turned up before, but they were ambiguous.

What the new data actually show

The streak was first noticed in a Hubble image in 2023, and interpreted then as a possible black hole wake on circumstantial grounds. Now the direct evidence is in. JWST’s NIRSpec integral field unit measured the motion of gas at the tip and found a velocity change of about 600 kilometres per second across roughly one kiloparsec. Fitted to a bow shock model, that discontinuity implies the black hole is travelling at about 954 kilometres per second, with an uncertainty of a little over 100 either way, on a path tilted around 29 degrees out of the plane of the sky and toward us.

Emission line ratios point the same way, matching what fast radiative shocks produce. The paper describes the case for a supersonic bow shock as very strong, bordering on overwhelming.

The mass is harder to pin down, because none of the signatures come from the black hole directly. An energy-conservation argument sets a floor of at least ten million solar masses. A separate estimate, based on the stellar mass of the galaxy it apparently left, points closer to twenty million.

What is new, and what was already known

In 2023, the discovery gave a streak and a hypothesis. What 2026 adds is the measurement that separates a moving object from the alternatives. Deep new Hubble imaging shows the emission dropping by a factor of more than forty at the tip, which is difficult to square with the two competing readings that had been proposed: an edge-on galaxy seen in profile, or a galaxy being torn apart. Neither produces a clean shock front at one end.

To their credit, the authors also correct parts of their earlier paper. A proposed second wake on the opposite side of the galaxy is not there in the deeper data. Their original reading of the velocity curve along the streak was wrong, and is now explained as gas cooling and mixing behind the shock rather than the object’s gravity tugging at the surrounding medium.

RBH-1 is not the Cosmic Owl

Some of the coverage has placed RBH-1 inside a system called the Cosmic Owl. In our reading of the paper, that is a mix-up worth heading off. The Cosmic Owl, also known as the infinity galaxy, appears in the introduction only as a contrasting case. It holds three active black holes, and they argue that its unusual central one most likely formed in place rather than being ejected. RBH-1 is a separate object entirely: the 62-kiloparsec feature beside a different galaxy. They get conflated because van Dokkum’s name sits on each, and both involve displaced black holes. They are not the same discovery.

Where the newborn stars come from

The wake is not the black hole manufacturing stars. As the object moves through the tenuous gas around its former host, it drives a shock and drags a cooling column of gas behind it. In that turbulent trail, gas from the surrounding medium mixes in, cools, and in places becomes dense enough to form stars where it sits.

That mechanism is plausible but not fully closed. On the paper’s own accounting, the gas swept up and entrained falls short of the observed stellar mass by a factor of several, unless the trail is converting gas into stars at rates seen only in the most extreme starbursts, or unless it is producing an unusually top-heavy mix of stars. The trail of young stars is real. Exactly how it built so many of them is not yet settled.

What the paper does not show, and what to watch

Every signature described here comes from the shocked gas, not from the black hole. The object itself remains unseen. There is a faint ultraviolet knot near the position the model predicts for it, which they flag as a possible glimpse of gas near the object itself, and are careful to call tentative.

How it escaped is also unresolved, though the paper leans one way. A black hole can be thrown out either by a three-body slingshot or by the recoil from a merger that radiates gravitational waves unevenly. Its escaped mass looks close to what the former host’s central black hole should weigh. That similarity points the authors toward recoil. In a slingshot, it is usually the lightest of three black holes that gets ejected, so the masses would not be expected to match.

A more consequential question is whether RBH-1 is one of many. Wakes like this may be a generic marker of ejected black holes, and if so, counting them would constrain how often galaxies fling their central black holes out. It is thin enough that ground-based surveys tend to smear it away, which is why the authors point to wide-field space telescopes, Euclid and the Nancy Grace Roman Space Telescope, as the instruments most likely to turn up a second example.