Researchers at East Texas A&M University have tested whether AI can tell from a drone's flight records that its GPS signal is being spoofed, DroneLife reports. Their study was accepted for WATCH 2026, a workshop held with the ACM Conference on Computer and Communications Security in The Hague.
The models ignore GPS data. Like dead reckoning, they check how fast the drone moves, in which direction and for how long, and look for movements that do not fit the reported position.
The team used a public set of 96 real flights, 54 of them flown under controlled spoofing attacks. It compared eight AI models. Models already trained on numerical time-series data did best. Given 10 consecutive readings, a model called Chronos scored 97.5% on a measure that weighs caught attacks against false alarms.
The study looked only at recorded flights. The university says the next step is to run the checks while the drone is flying.
“We don’t refer to any GPS-related attributes,” said Jinoh Kim, a professor at the university.
Background
East Texas A&M University
University
A public research university in Commerce, Texas, formerly called Texas A&M University–Commerce.
Jinoh Kim
Professor at East Texas A&M University
Professor in the Department of Computer Science and Information Systems at East Texas A&M University.



