AI Is Now Tracking Illegal Fishing in Real Time From Space — What the New Satellite System Can See That Humans Cannot

For most of human history, the ocean kept its secrets well. A fishing vessel a hundred miles offshore, operating at night, with its lights and transponders switched off, was essentially invisible to anyone trying to enforce a law against it. Coast guards could patrol, satellites could photograph, but the sheer scale of the open ocean meant that illegal fishing has always thrived in the simple, brutal mathematics of too much water and too few eyes. That mathematics is now changing. A growing system of satellites, radar, and artificial intelligence, led primarily by the nonprofit organization Global Fishing Watch, is making it possible to see vessels that have spent decades successfully hiding from exactly this kind of scrutiny, and the implications for the health of the world's oceans are significant.

How the System Actually Works

Global Fishing Watch's technology is built on a foundation of combining several different data sources that, individually, each have serious blind spots, but together produce a far more complete picture of what is actually happening at sea. The first layer is the automatic identification system, known as AIS, a tracking technology that many commercial vessels broadcast publicly every few seconds, originally designed to help ships avoid colliding with one another. The second layer is vessel monitoring systems, or VMS, a similar form of tracking that has historically been restricted to government fisheries regulators rather than made publicly available. Both systems are useful, but both share the same fundamental weakness: a vessel can simply switch its tracking device off, and once it does, it effectively disappears from these data streams entirely.

This is where satellite imagery and machine learning enter the picture, and where the system becomes genuinely capable of seeing what these traditional tracking methods cannot. Global Fishing Watch combines AIS and VMS tracking with radar imagery, optical imagery, and nighttime lighting data captured from satellites, then uses artificial intelligence to analyze that imagery and identify vessels that never appear in any tracking broadcast at all. The radar technology involved, known as synthetic aperture radar, is particularly powerful because it can detect a solid metal object like a ship's hull regardless of weather, cloud cover, or whether the vessel is broadcasting any signal whatsoever. Machine learning models trained on this combined imagery can then estimate a detected vessel's length, classify whether its movement pattern resembles fishing activity, and flag it for further investigation, all without that vessel ever knowing it has been spotted.

What This System Can See That Human Patrols Cannot

The scale of what this approach reveals is genuinely striking. In one major Global Fishing Watch study analyzing radar detections inside marine protected areas, researchers found that roughly 75 percent of fishing vessels detected by satellite radar were not being tracked by AIS at all, meaning these vessels were either deliberately concealing their location or operating in a manner that fell entirely outside the visibility of conventional tracking methods. A separate, larger research effort, conducted in partnership with researchers from the University of Wisconsin-Madison, Duke University, UC Santa Barbara and SkyTruth, analyzed roughly 2 million gigabytes of satellite imagery captured between 2017 and 2021, mapping vessel traffic and offshore industrial infrastructure across coastal waters on six continents where more than three-quarters of all industrial ocean activity is concentrated.

No human patrol boat or coast guard vessel could conceivably replicate this kind of coverage. A patrol boat can inspect a handful of vessels within sight of its own position on a given day. A satellite-and-AI system can analyze imagery spanning enormous stretches of ocean continuously, flagging suspicious vessels for human investigators to follow up on rather than requiring those investigators to search blindly across thousands of square miles of open water. This shift, from active searching to AI-assisted triage, is precisely what makes the technology so valuable to chronically under-resourced fisheries enforcement agencies, particularly in developing coastal nations that could never afford the kind of large patrol fleet that comprehensive ocean surveillance would otherwise require.

The technology has also proven effective at evaluating whether existing conservation measures are actually working. A study using this same radar-and-AI approach examined 1,380 marine protected areas around the world and found that 78.5 percent experienced no detectable commercial fishing activity at all, with illegal fishing detected in the remainder typically averaging less than 24 hours per year in 82 percent of cases. That kind of granular, continuously updated assessment was simply unavailable to ocean conservation researchers before this combination of satellite imagery and machine learning became practical.

How This Technology Connects to the Mombasa Declaration

This growing surveillance capability is not developing in isolation from international policy. In June 2026, fifteen nations spanning Africa, Asia, Europe, the Caribbean and the Pacific signed the Mombasa Declaration at the 11th Our Ocean Conference in Kenya, committing signatories to modernize their vessel registries, publish fishing authorizations, and strengthen the sharing of information about vessel ownership and licensing. That declaration explicitly relies on the kind of technological visibility Global Fishing Watch and similar organizations provide, since transparency commitments are only meaningful if governments actually have the tools to verify what vessels are doing once those commitments are made on paper.

During the same conference, Global Fishing Watch announced a new partnership with the Minderoo Foundation to build what organizers describe as the first global map capturing both large industrial fishing fleets and the millions of smaller artisanal vessels that have historically been excluded from these tracking systems entirely, alongside individual country partnerships, including new arrangements with Madagascar and Panama to share vessel tracking data more openly. The relationship between the Mombasa Declaration's political commitments and Global Fishing Watch's technical capabilities is, in effect, a partnership between policy and enforcement infrastructure, since one without the other accomplishes relatively little. A government can commit to greater transparency, but without the satellite and AI tools to actually detect violations, that commitment has limited practical force. Conversely, even the most sophisticated detection technology cannot enforce laws that governments are unwilling to act on once violations are identified.

Where Illegal Fishing Is Concentrated Today

Understanding where this technology matters most requires understanding where illegal, unreported and unregulated fishing is actually concentrated, and the geographic pattern is sharp and consistent across multiple independent research efforts. West Africa is widely considered the world's single worst hotspot for this kind of activity, losing an estimated $9.4 billion annually to illegal fishing, driven substantially by distant-water fishing fleets, with China's fleet, the largest in the world, identified repeatedly as the most significant offender according to the IUU Fishing Risk Index. Southeast Asia represents the other major global hotspot, where rapid economic growth, surging regional demand for seafood, and contested maritime boundaries combine to create conditions that are particularly difficult to police, with regional monitoring centers tracking hundreds of illegal fishing incidents annually across the wider Indian Ocean region.

East African nations face a related but distinct version of this problem. Research from the World Wildlife Fund found that Kenya, Madagascar, Mozambique, South Africa and Tanzania collectively lost up to $142.8 million annually between 2015 and 2021 due specifically to illegal shrimp and tuna fishing, with the affected catch including dozens of shark and ray species alongside a wide variety of reef fish. Globally, illegal fishing is now understood to account for roughly one-fifth of total reported fisheries catches, worth as much as $23.5 billion annually, making it, according to research compiled by the Financial Transparency Coalition, the third most lucrative natural resource crime in the world after illegal timber and illegal mining.

What AI-Driven Enforcement Means for Ocean Health

The long-term significance of this technology extends well beyond catching individual violators, important as that enforcement function is. Illegal fishing is a major contributor to the depletion of global fish stocks, with the United Nations estimating that more than 90 percent of global fisheries stocks are now fully exploited, overexploited, or depleted, a trajectory that satellite-and-AI detection technology offers a genuine, if partial, tool to help reverse. When marine protected areas can be continuously monitored rather than occasionally patrolled, the conservation value of those protected zones increases substantially, since fish populations need sustained, reliable protection over years to recover meaningfully rather than intermittent enforcement that poachers can learn to work around.

There is also a transparency dimension to this technology that extends beyond direct enforcement. Because much of Global Fishing Watch's vessel tracking data and resulting maps are made freely and publicly available, researchers, journalists, advocacy organizations and policymakers around the world can independently verify claims about fishing activity in specific waters, rather than depending entirely on the self-reported figures of national fisheries agencies or the fishing industry itself. That kind of independent, satellite-verified visibility makes it considerably harder for illegal fishing operations to persist quietly for years, as historically often happened, particularly in remote waters far from any country's coastline where enforcement capacity has traditionally been weakest.

None of this means the technology has solved the underlying problem. Detecting a vessel and successfully prosecuting the people responsible for illegal fishing remain two very different challenges, and many of the countries facing the worst IUU fishing problems still lack the patrol vessels, legal frameworks, or diplomatic leverage needed to act decisively once a violation has been identified by satellite. Complex ownership structures and flags of convenience continue to shield the people who actually profit from illegal fishing operations, often leaving fines and prosecutions to fall on vessel captains and crew rather than the corporate or individual owners truly responsible. Closing that gap between detection and accountability remains, by the admission of organizations working in this space, one of the central unresolved challenges of ocean governance today.

The Bottom Line

The combination of satellites, radar and artificial intelligence has fundamentally changed what is knowable about activity on the open ocean, turning what was once one of the most opaque environments on Earth into something approaching continuous, verifiable surveillance. Vessels that spent decades successfully evading detection by simply switching off a transponder are now visible through radar and optical imagery that no amount of darkness or radio silence can fully conceal. Whether that visibility translates into meaningfully healthier oceans depends on what happens after a vessel is detected, on whether governments follow through on commitments like the Mombasa Declaration, on whether enforcement agencies in the world's most affected regions receive the resources and legal tools to act on the information this technology provides, and on whether the people who actually profit from illegal fishing can finally be held accountable rather than the crews who simply operate the boats. The ocean is no longer too big to watch. Whether the world chooses to act on what it now sees is the question this new visibility has only just begun to answer.

*This article is for informational purposes only. Data and research findings are sourced from Global Fishing Watch, Space.com, the Global Investigative Journalism Network, the IUU Fishing Risk Index, the Financial Transparency Coalition, and the Stimson Center.*