February 25, 2026 — The biggest challenge with studying the world’s fish population is simple: fish live underwater.
You can’t easily easily spot them from land or even from space with satellites so getting a sense for where they are when, where they are going, and how many of them are left is really hard.
But one non-profit, Global Fishing Watch, is now trying a cutting-edge technique that they hope will help better track the movements of the world’s fish population: they are tracking the world’s fishermen.
The group leverages positioning data from Automatic Identification System (AIS) transponders that are required on large vessels, primarily to prevent collisions.
But taking that data and adding in a powerful dose of Artificial Intelligence, the group’s chief scientist David Kroodsma claims they are able to paint a global picture of fishing fleets.
“What we do is we take all the GPS positions of all the boats in the world. You’re talking about a database of 100 or well over 100 bazillion, 100 billion, GPS positions. And we use machine learning to determine what types of boat they are, what they’re doing, if they’re fishing boats, when they’re fishing,” Kroodsma explains.
And if a ship flips off the transponder for any reason, a maneuver known as going dark, Kroodsma and his fellow trackers double down.
“Millions of gigabytes of satellite imagery to and apply AI to that to detect vessels and then determine which vessels are broadcasting which ones or not, to come up with that larger picture of where activity on the ocean is happening,” he adds.