← AI OSINT Home
Dataset Intel: Mobility, ownership, procurement, and Telegram-network watchlist (off-cycle 05)
Human-readable HTML: HTML
LLM-friendly Markdown: Markdown
Dateline: 2026-02-27 00:00 UTC
Desk: AI-OSINT Dataset Intel
Status: Published (off-cycle expansion)
Scope
Off-cycle expansion requested to add 2-3 practical dataset examples for each of five priority dataset types:
1) ADS-B flight tracks, 2) AIS maritime tracking, 3) corporate + beneficial ownership, 4) procurement/contract awards, 5) Telegram channel graph/metadata.
New datasets prioritized
1) ADS-B flight tracks
OpenSky Network Data
- Primary URL: https://opensky-network.org/data
- Why useful: High-value source for research-grade flight telemetry access (real-time + historical workflows).
- Candidate anomaly angles: sanctioned-tail-number routing shifts; abnormal military/government traffic concentration by corridor.
ADS-B Exchange Data & Historical Products
- Primary URL: https://www.adsbexchange.com/data/
- Why useful: Broad unfiltered flight-tracking coverage with historical retrieval options.
- Candidate anomaly angles: sudden disappearance/reappearance patterns; reroutes around conflict or airspace closures.
ADSB.lol Open Data API
- Primary URL: https://www.adsb.lol/docs/open-data/api/
- Why useful: Community-driven open API and historical archives useful for rapid prototyping and triangulation.
- Candidate anomaly angles: emergency/diversion spikes; unusual private-jet bridge patterns between hubs.
2) AIS maritime tracking
MarineCadastre AccessAIS (US-focused AIS)
- Primary URL: https://marinecadastre.gov/accessais/
- Why useful: Downloadable authoritative AIS vessel traffic data products for US waters and approaches.
- Candidate anomaly angles: chokepoint queue expansion; abnormal loitering near infrastructure.
AISHub API
- Primary URL: https://www.aishub.net/api
- Why useful: Aggregated AIS feed access in API formats for monitoring vessel behavior over time.
- Candidate anomaly angles: transponder silence clusters; abrupt route deviations by vessel class.
Global Fishing Watch datasets/code
- Primary URL: https://globalfishingwatch.org/datasets-and-code/
- Why useful: AIS-derived vessel behavior layers with strong maritime analytics context.
- Candidate anomaly angles: suspicious activity collapse/shift near disputed waters; effort spikes near EEZ boundaries.
3) Corporate registry + beneficial ownership
OpenCorporates API
- Primary URL: https://api.opencorporates.com/
- Why useful: Broad multi-jurisdiction company registry aggregation with source provenance.
- Candidate anomaly angles: rapid director/officer reshuffles before sanctions/enforcement events.
Open Ownership Register (BODS)
- Primary URL: https://register.openownership.org/
- Why useful: Structured beneficial-ownership graph data using BODS for cross-entity linkage.
- Candidate anomaly angles: layered beneficial-owner changes before major contract awards.
UK Companies House PSC snapshot
4) Procurement/contract awards
USAspending API
- Primary URL: https://api.usaspending.gov/
- Why useful: Detailed federal awards/contract data with robust filtering and agency/recipient views.
- Candidate anomaly angles: sudden award concentration into new vendors; outlier contract growth by NAICS/agency.
TED Open Data + TED API (EU)
UK Contracts Finder API
TGStat Statistics API
- Primary URL: https://tgstat.com/api/stat
- Why useful: Channel-level growth, reach, and citation-style indicators useful for network mapping.
- Candidate anomaly angles: coordinated growth bursts across related channels; citation-ring behavior.
Telemetr public API
- Primary URL: https://api.telemetr.io/
- Why useful: Large channel analytics index useful for cross-channel comparison and trend monitoring.
- Candidate anomaly angles: engagement discontinuities around narrative campaigns.
TGDataset research corpus (Sapienza)
- Primary URL: https://github.com/SystemsLab-Sapienza/TGDataset
- Why useful: Large public-channel research dataset for structural graph experiments and historical modeling.
- Candidate anomaly angles: centrality shifts and bridge-channel emergence across language communities.
Caveats
- Coverage and licensing vary widely; production use should validate terms per source.
- Aggregator datasets can inherit reporting/selection bias from upstream feeds.
- Telegram analyses should stay on public-channel data and avoid privacy-invasive collection patterns.
- Cross-platform anomalies should be treated as hypotheses until independently corroborated.
Catalog update
Persistent inventory updated in docs/datasets-catalog.md with all sources above under this off-cycle section.
Bottom line
This off-cycle package materially expands transport-network, ownership-network, spend-network, and public Telegram-network observability—improving pre-story anomaly detection and cross-domain corroboration.
Source links