Methodology
How TrackWanted works
TrackWanted is an open-source intelligence reader. It collects material that is already public, summarises it with citations, and is explicit about the difference between something that was reported and something that was modelled.
Where the information comes from
- Wanted notices and designations — Interpol red notices, national agency listings, sanctions designations and court records.
- Aircraft identity — the FAA civil aircraft registry as the primary source, with Planespotters as a secondary cross-check. Every tail number in the app links to both.
- Open reporting — a live sweep across news indexes and the GDELT global event database, run with name variants so regional-language coverage is not missed.
- Public financial records — the ICIJ Offshore Leaks database, open company registries and public blockchain explorers.
What the AI layer actually does
Language models are used for reading, not for inventing. They extract structured facts — who, where, when, how much — from documents the system has already retrieved, and anything that cannot be traced back to a retrieved source is discarded before it reaches the screen. Every brief item carries the link it came from. If nothing credible was published, the app says so instead of filling the gap.
What is modelled, not measured
The globe is a visualisation. Aircraft positions are simulated along realistic routes and schedules to illustrate movement patterns; they are not a live air-traffic feed and must never be used for navigation or operational decisions. Carbon figures are estimates derived from published fuel-burn rates and great-circle distances. Fugitive location rings show a probability area around the last publicly reported region, not a fix.
Privacy and responsible use
TrackWanted collects nothing about its visitors: no accounts, no tracking pixels, no third-party analytics, and referrers are stripped on outbound links. It publishes only information about public figures, public aircraft registrations and published legal notices. Wanted status reflects what an authority has published and is not a determination of guilt. The platform is intended for research, journalism and public accountability.
Corrections and takedowns
Public records change: notices are withdrawn, registrations are transferred, cases are resolved. When a source updates, the board updates with it. If you believe an entry misstates the public record, the contact page reaches the operator directly and corrections are made against the same published sources the entry came from.
Frequently asked questions
Does TrackWanted use any private or leaked data?
No private databases and nothing obtained by intrusion. The system reads published notices, public registries, court and sanctions records, and datasets that are openly distributed — including journalistic leak databases such as ICIJ Offshore Leaks that are published for public research.
Can I trust the AI-written summaries?
Each summary is constrained to documents the system has already retrieved, and every item carries the link it came from so you can check the original. Content that cannot be traced to a retrieved source is discarded automatically rather than shown.
How often is the data refreshed?
The wanted board and money trail refresh from open reporting twice a day. The aircraft registry index is rebuilt weekly, and air-quality readings on the carbon board update every ten minutes.
Who is TrackWanted for?
Journalists, researchers, analysts and curious readers. It is a reading and research aid for public accountability — not an operational tool, a navigation source, or a way to locate any individual in real time.