The driving process is called streaming instability — the tendency of solid particles in a disk of gas to drag against the gas and spontaneously clump together. It is the same mechanism implicated in planet formation in ordinary solar systems.

The numbers involved are striking. The researchers estimate that roughly a million planets could form during a single active episode of an AGN, and one such episode typically lasts about 2 million years. A black hole can cycle through many of these feeding periods over billions of years, so the cumulative count of planets formed could be immense.

Some of these worlds would face a grim end, though. The black hole's gravity could drag planets inward, where temperatures would be hot enough to disintegrate them — ironically producing yet more dust for future planet formation.

Could we ever detect them?

Not yet, and not easily. The paper outlines why the standard planet-hunting toolbox falls short:

• Radial velocity — detecting the wobble a planet induces in its star — fails because black holes are simply too massive to wobble measurably.
• The transit method, which watches for a planet passing in front of its star, might work in principle. But even Jupiter-mass worlds would be vanishingly small against the accretion disk, the only available background light source, since a black hole itself emits no light.
• Gravitational microlensing, in which a planet's mass distorts the surrounding spacetime, could also work. But it would require several planets around the same black hole plus a favorable alignment, and the noisy environment around a supermassive black hole would make a planet's signal hard to isolate.

The researchers say confirming these objects will require new detection techniques altogether.

Are they really planets?

Even the peer-review team pushed back on calling these objects planets, given how different their birth environment is from a star's disk. Lyra suggested the term "blanets" — black hole planets. The name has not caught on in the field, so the researchers currently settle for the more cautious label "planetary-mass objects."

Astronomer Phil Plait voiced the general surprise on social media: "Pretty much the last place in the universe I'd expect to find planets is in the incredibly violent and radiation-laden accretion disks around supermassive black holes. And yet… maybe?"

Could these worlds have moons — or life?

Moons are unlikely in the traditional sense. The researchers say these objects would more probably form as binaries — two bodies of similar mass orbiting each other, like binary asteroids found in our solar system. Angular momentum during formation would favor such pairings, much as it does in the Kuiper Belt, where astronomers keep finding more binary objects.

As for life: nobody knows, but it is not flatly impossible. The simulations suggest these worlds would have an energy source even without starlight — the hot gas of the accretion disk itself.

The caveat bears repeating: these results come from simulations, not observations. The study establishes that black hole planets are physically plausible, not that they exist. Confirming them, or ruling them out, will fall to future instruments and methods that have yet to be invented.