Global Fishing Watch upgrades AI to sharpen fishing estimates worldwide

Global Fishing Watch has rolled out a major update to its global ocean mapping tools, using new AI models and geographic data to better estimate fishing activity and classify vessels. The changes, available now across its map, APIs and downloadable products, are meant to give governments, researchers and civil society a clearer view of what is happening at sea.

Why it matters: - Global Fishing Watch’s upgraded data can change how governments, researchers and ocean advocates interpret fishing activity, marine protected areas and fleet behavior. - Better vessel identification and fewer false positives should help users focus on activity that is more likely to matter for enforcement, science and ocean management. - The update may lower some historical and current fishing estimates, which affects analysis across policy, research and public transparency work.

What happened: - Global Fishing Watch launched a major technology upgrade to its global ocean mapping system on Oct. 21, 2026. - The update is live across the organization’s public map, APIs and downloadable data products. - The release combines new artificial intelligence models with improved geographic datasets to track fishing activity and identify vessel types. - The upgrade covers data from 2012 to the present.

The details: - The new fishing model uses more than 1,000 expert-reviewed examples of fishing for training. - The model also uses a new algorithm architecture plus updated anchorages and geographic boundaries to determine whether a vessel is actually fishing. - The changes are designed to reduce false positives from vessels slowing near ports, sitting at anchor or transiting without fishing. - The upgraded system improves classification of fishing vessels versus cargo ships, tankers and other vessel types. - Thousands of vessels have been reclassified more accurately between fishing and non-fishing categories. - Detection of fishing gear and drifting buoys in the water has also improved, reducing another source of false positives. - Global Fishing Watch added more than 5,300 new anchorages across 13 countries. - The new anchorages improve distance-to-port calculations and help eliminate erroneous port visits. - A new unified spatial grid standardizes information such as distance from shore and distance from port. - Higher-resolution bathymetry improves estimates of ocean depth. - Users can now analyze fishing activity alongside updated reference layers, including FAO Major Fishing Areas for Statistical Purposes, General Fisheries Commission for the Mediterranean Fisheries Restricted Areas and the ProtectedSeas marine protected area database.

Between the lines: - This is the first major structural update to Global Fishing Watch’s core AI models for estimating apparent fishing activity and classifying vessels since the organization was founded. - The retrospective model changes mean some historical estimates and vessel classifications will shift, which is important for anyone comparing older and newer analyses. - Global Fishing Watch will keep both the previous and upgraded datasets available for a limited time so users can compare outputs. - The organization is inviting feedback through its reporting tools, and local partner knowledge will continue to inform model refinement. - The upgraded models still have limits, including possible misclassification from model bias or gaps in source data. - Global Fishing Watch recommends using its data with official vessel registries, local intelligence and original tracking information before making legal or regulatory decisions.

What’s next: - Users will be able to compare old and new datasets for a limited period to understand how the upgrade affects their work. - Continued feedback from users and partners is expected to shape future refinements to the models. - The organization says the data will keep improving as the technology and underlying understanding of ocean activity evolve.

The bottom line: - Global Fishing Watch is betting that better AI and cleaner geospatial data will make its ocean monitoring tools more accurate, more transparent and more useful for real-world decision-making.