NASA’s AI Tool May Have Uncovered Thousands of New Alien Planets

What happened:
NASA scientists say an artificial intelligence system may have identified thousands of previously unknown exoplanets, dramatically expanding the list of worlds orbiting distant stars.

Why it matters now:
The discovery comes as space agencies prepare for the Nancy Grace Roman Space Telescope, which is expected to detect tens of thousands of new planets. AI is emerging as a critical tool to process the overwhelming volume of space data already collected.

---

AI digs deeper into space data

NASA confirmed in late 2025 that the number of officially confirmed exoplanets crossed 6,000, with another 8,000 awaiting verification. However, researchers believe the real number is far higher.

A major reason is ExoMiner, an AI-driven software developed at NASA’s Ames Research Center. Designed to re-analyse existing telescope data, the system scans for subtle signals that human researchers might miss.

Using data from the retired Kepler Space Telescope and the still-active TESS (Transiting Exoplanet Survey Satellite), ExoMiner has already shown strong results. An early version identified 370 previously unknown planets hidden in Kepler data.

---

ExoMiner++ flags nearly 7,000 new candidates

A newer version, ExoMiner++, detailed recently in The Astronomical Journal, has now flagged nearly 7,000 potential exoplanets in TESS data during its initial run alone.

While these candidates are not yet confirmed, the AI has provided astronomers with precise targets for follow-up observations, significantly speeding up the discovery process.

Kepler and TESS together account for nearly half of all known exoplanets, meaning large portions of their archived data still hold untapped discoveries.

---

Open-source science accelerates discovery

NASA has made ExoMiner++ freely available on GitHub, allowing researchers, students, and citizen scientists worldwide to analyse real space data.

“Open-source software like ExoMiner accelerates scientific discovery,” said Kevin Murphy, NASA’s chief science data officer. “Sharing tools lets others replicate results and push research further.”

The software connects directly with TESS’s public data archive, opening planet-hunting to the global scientific community.

---

Looking ahead: Roman Telescope and beyond

NASA hopes ExoMiner and similar AI tools will play a key role in analysing data from the Nancy Grace Roman Space Telescope, scheduled for launch by May 2027.

If successful, the AI could help sustain the explosive growth in planet discoveries that began in 1995 with the detection of 51 Pegasi b, the first exoplanet found orbiting a Sun-like star.

The next frontier, scientists say, is developing AI capable of identifying which of these distant worlds might support life — a challenge that could redefine humanity’s place in the universe.

What to watch next:
Follow-up observations to confirm ExoMiner++ planet candidates — and whether future AI tools can narrow the search for habitable worlds.