The Ocean World That Won't Answer: Two Years Inside the K2-18b Alien-Life Debate
Some 124 light-years away, in the direction of the constellation Leo, sits a planet that has spent two years as the most exciting and most argued-about object in astronomy. K2-18b is a sub-Neptune roughly nine times Earth's mass, circling a dim red dwarf star inside its habitable zone. Depending on which paper you trust, it either has a warm ocean that might support life, or a thick chemical haze and nothing worth writing home about.
Both camps have data from the James Webb Space Telescope. Both camps have math. Neither has backed down.
The trouble started in April 2025, when Nikku Madhusudhan's team at Cambridge announced JWST observations that appeared to show dimethyl sulfide, or DMS, in the planet's atmosphere. On Earth, DMS is produced almost entirely by marine plankton — it's part of what gives the ocean its smell.
Finding it on a distant "Hycean" world (Madhusudhan's own term for an ocean planet wrapped in hydrogen) was cautiously framed as the closest thing to a biosignature ever detected outside our solar system. NASA never called it a discovery of life. Plenty of headlines did it for them anyway.
What followed was a very public demonstration of how science actually works — which is to say, messily. Independent teams went after the statistics, and one paper, bluntly titled K2-18b Does Not Meet the Standards of Evidence for Life, argued the detection simply didn't clear the bar. A joint reanalysis of three separate JWST instruments' data agreed.
Then came the heavyweight rebuttal. A NASA-led study out of the Jet Propulsion Laboratory — with Madhusudhan himself among the authors — combined four new JWST observations with the original data and had three independent teams run the analysis.
The verdict: K2-18b really is rich in water, with a complex and interesting atmosphere, but there is no conclusive DMS. The earlier signal was most likely "red noise" — instrumental static dressed up as chemistry. A planet that might have smelled like the ocean, reduced to a rounding error.
That would have been a tidy, if deflating, place to end the story. Instead, a completely different team pointed two of Earth's most sensitive radio telescope arrays at the same planet and asked a different question entirely: is anyone there transmitting?
In February 2026, a team led by Chenoa Tremblay used the Very Large Array in New Mexico and the MeerKAT array in South Africa to listen for artificial radio signals from the K2-18 system. They watched a huge stretch of the radio spectrum across enough sessions to cover the planet's full 33-day orbit — one entire K2-18b year.
Notably, Madhusudhan is a co-author on this paper too. The same scientist who helped make K2-18b famous for a maybe-smell of ocean life spent the following year listening for it to say something out loud.
It did not. The raw data produced millions of signal hits, which is normal — satellites, human technology, and the instruments themselves generate enormous noise. The team's real contribution was the filtering pipeline built to cut through it, testing each candidate against the physics of the system's orbit and the geometry of the telescope beams.
Every one of those millions of hits failed the full filter. Zero technosignatures, down to a sensitivity of roughly a trillion watts — which sounds enormous until you consider that's what it takes to catch a deliberate, steady broadcast from 124 light-years away.
It was the first search of its kind ever run on a candidate ocean world. And the team is fairly explicit that the real product isn't the null result — it's the reusable method for searching the next dozen planets like it.
There's something almost soothing in the contrast between the two efforts. The biosignature hunt is inherently messy: you stare at bumps in a noisy spectrum and argue over whether they mean plankton or nothing. The radio search is comparatively clean — either a signal behaves like deliberate engineering or it doesn't, and K2-18b failed one test ambiguously while passing the other cleanly, in the "nobody's home" direction.
None of this settles what K2-18b actually is. The water-rich atmosphere is real, and the global-ocean idea is still standing — neither confirmed nor killed. The DMS claim is functionally dead for now, pending observations nobody has made yet.
And the radio silence, however carefully measured, rules out only one very specific kind of neighbor: one broadcasting continuously, loudly enough for 2026-era instruments to notice. It says nothing about oceans, or microbes, or anything quieter than a radio tower.
K2-18b isn't the only target of this kind of scrutiny lately, either. Just days ago, a separate team published JWST follow-up observations of two candidate Dyson spheres — hypothetical megastructures built to capture a star's entire energy output. Nothing there has been confirmed, and probably won't be; these searches almost never end that way.
Taken together, though, it's a fair snapshot of where the hunt for extraterrestrial life sits in mid-2026: less "have we found aliens," more "we've gotten remarkably good at saying, with real statistical teeth, exactly what we haven't found." For now, K2-18b remains what it's been for two years — the exoplanet everyone wants to be the one. Water-rich, atmospherically strange, plausibly ocean-covered, and, as far as the world's best radio telescopes can tell, not saying a word.