Scientists Just Found 10,000 Possible New Planets
Humanity currently knows of about 6,286 confirmed planets outside our solar system. A new study just added more than 10,000 candidates to that list in one shot — which would more than double everything we've found in the three decades since the first exoplanet was discovered.
The find came from a team of researchers who did something surprisingly simple — they went back and looked at old data nobody had fully examined before.
NASA's Transiting Exoplanet Survey Satellite, known as TESS, has been hunting planets since 2018 and has confirmed nearly 900 of them on its own. The way it works is by watching for starlight to dim slightly — when a planet passes in front of its star from TESS's vantage point, the star gets a little darker for a brief period. That transit is how you know something is there.
The problem is that TESS has been observing an enormous number of stars and researchers have mostly focused on the brightest ones — the ones easiest to analyze. About 83 million fainter stars in the 2018 TESS dataset had been largely ignored.
So this team fed that overlooked data into a machine learning system and let AI comb through it looking for the telltale dimming patterns that signal a planet transit. What came back was 10,091 possible planet-like objects that nobody had spotted before.
What "Candidates" Actually Means
These aren't confirmed planets yet. They're candidates — objects that look like planets based on the light data but need additional verification before they can be officially added to the confirmed count. Some of them will turn out to be something else when examined more closely. But the sheer volume of the find, and the technique used to make it, is what has astronomers excited.
The study has been published on the research repository ArXiv and is awaiting peer review — so it hasn't gone through the full scientific vetting process yet. But the methodology is already being described as a significant leap forward in how researchers hunt for planets. If AI can find 10,000 candidates in one year of overlooked data from one satellite, the implications for what future analysis might turn up are significant.
Why This Actually Matters
Since the first exoplanet was confirmed in 1995, researchers have been building toward one specific goal — finding a planet that resembles Earth closely enough to potentially support life. None of the 6,286 confirmed exoplanets found so far has fully cleared that bar.
The closest anyone has gotten recently was exoplanet K2-18b, which made headlines in April 2025 when a team of astronomers claimed to have found in its atmosphere what they described as the strongest evidence yet for life existing beyond Earth. Other scientists pushed back on those findings pretty quickly and the scientific community remains divided on what the data actually shows.
More candidates mean more chances. More planets to examine, more atmospheres to analyze, more data to feed back into the search.
What's Coming Next
NASA is planning to launch the Nancy Grace Roman Space Telescope in 2026 — a significantly more powerful instrument designed specifically to discover thousands of new exoplanets through a technique called gravitational microlensing, which detects planets by watching how their gravity bends light from distant stars. Combined with AI analysis of existing data, this new telescope could accelerate the pace of discovery dramatically.
The search is also happening closer to home. NASA spacecraft are already targeting Europa — a moon of Jupiter — with a probe due to arrive in 2030 to search beneath its icy surface for water and signs of habitability. Saturn's moon Enceladus is also considered a serious candidate for harboring life.
Ten thousand new worlds to look at. One Earth-like planet still waiting to be found. The search just got a lot bigger.
Curious for more stories that keep you informed and entertained? From the latest headlines to everyday insights, YourLifeBuzz has more to explore. Dive into what’s next.