
NVIDIA is working with the SETI Institute and Breakthrough Listen to use agentic artificial intelligence in the search for extraterrestrial signals at California's Allen Telescope Array.
Announced on 30 October 2026, the system combines telescope operations, real-time data processing and AI analysis, although the project has not reported any confirmed detection of extraterrestrial intelligence.
The initiative addresses a longstanding challenge in radio astronomy. Telescopes collect vast quantities of data, and researchers must identify potentially interesting signals among the information captured during observations.
Automating parts of that process could help scientists investigate signals sooner and make more efficient use of telescope time.
How NVIDIA AI Is Changing the Search for Extraterrestrial Signals
The new system allows AI agents to coordinate several stages of the observation process through natural-language instructions. Researchers can use a single request to direct telescope operations, process incoming data and apply AI models while observations are under way.
NVIDIA's Holoscan platform processes telescope data as it arrives, allowing researchers to analyse signals in near real time.
BREAKING: NVIDIA put AI agents in charge of a radio telescope hunting for alien signals.
— Guillermo Flor (@guilleflorvs) November 1, 2026
At the Allen Telescope Array in Northern California, the SETI Institute and Breakthrough Listen now run the instrument in plain English:
1. Agents steer the 42 antennas
2. They calibrate… pic.twitter.com/KNr4l7dTcN
DAQIRI transfers sensor data directly into GPU memory, while CyberEther, developed by the SETI Institute, uses NVIDIA's CUDA-X libraries, including cuFFT and cuBLAS, for signal processing.
Nexus connects AI agents with the telescope's status, computing resources and processing pipelines, allowing researchers to monitor data loss, inspect live spectrograms and adjust signal processing through natural-language prompts.
NVIDIA says the system can handle up to 2.4 terabits of data per second using RTX PRO 6000 Blackwell GPUs.
By integrating telescope operations with real-time analysis, the technology aims to streamline research, although its time-saving benefits have not been independently established.
AI Detects a Pulsar, but Alien Signals Remain Unconfirmed
The project has demonstrated the system's ability to identify a pulsar in the Crab Nebula during real-time data processing, according to NVIDIA. A pulsar is a rapidly rotating stellar remnant that emits pulses of radiation, providing astronomers with a recognisable signal to study.
The latest development builds on AI-assisted pulsar detections reported by NVIDIA and the SETI Institute in 2025. Its focus is integrating agentic AI into telescope operations, data processing and computing-resource management, rather than achieving the first AI detection of a pulsar.
No extraterrestrial signal has been confirmed, and identifying astronomical signals does not establish the existence of alien intelligence.
Andrew Siemion, the SETI Institute's Bernard M. Oliver chair for SETI and principal investigator of Breakthrough Listen, has also discussed the possibility that extraterrestrial civilisations could have developed artificial intelligence of their own.
👻 Things that go BEEP in the night!
— The SETI Institute (@SETIInstitute) October 31, 2026
Could artificial intelligence help us find extraterrestrial intelligence?
At the SETI Institute's Allen Telescope Array, researchers are exploring how agentic AI and @nvidia technology can help automate telescope operations, process enormous…
'As our new agentic observatories reach out into the universe, what they find may have much more in common with themselves than with the humans who created them,' Siemion said. 'But this isn't something to fear. Rather, it's a breakthrough in translation.'
The possibility that extraterrestrial civilisations possess advanced AI remains speculative. NVIDIA's announcement refers to the idea that civilisations, if they exist, could be billions of years older than humanity, but it offers no evidence that such societies have been discovered.
The team is also evaluating open models, including NVIDIA Nemotron, for situations where data must remain local and results are needed quickly.
NVIDIA identifies potential applications in observatories that might eventually operate offline on the far side of the Moon, although this remains a prospective use rather than a confirmed deployment.
For now, the immediate goal is to give researchers more flexible tools for operating telescopes and analysing the data they collect. The search for extraterrestrial intelligence continues, with no confirmed discovery reported by the project.























