OONI
Network measurements and interference observations.
Visit the sourceVoidly / Observatory
A connection fails. The harder question is why. Voidly brings measurements, source records, and network context together so restrictions can be examined and cited.
Snapshot: . Read the coverage definitions and sources. Sample availability varies by country and time window.
The method
The observatory combines Voidly probe observations with external measurement projects. DNS, connection, TLS, and HTTP results describe where communication fails; routing and connectivity signals provide wider outage context.
A failure can also come from a misconfigured service, an unreliable network, or the measurement itself. Sources, dates, status, and confidence belong with the finding. An anomaly alone does not establish censorship.
Read the current methodologyThe evidence path
01 / Measurements
Keep the source, time, location scope, and tested service attached to each observation.
02 / Evidence
Group related observations, compare independent sources where available, and preserve uncertainty.
03 / Public record
Expose incident records, evidence links, country summaries, and machine-readable access.
Not every incident has multiple sources. The public record distinguishes suspected events, citable records, linked evidence, and independent corroboration. Inspect those categories.
Source foundation
Network measurements and interference observations.
Visit the sourceRemote measurements of internet censorship.
Visit the sourceConnectivity and routing signals for outage context.
Visit the sourceTest-list context for domains and categories.
Visit the sourceSource attribution describes data provenance. Each project has its own coverage, methodology, update schedule, and reuse terms.
Use the record
Start with the country or incident explorer. Move into JSON, downloadable datasets, or a developer integration when the work calls for it.
Keyless JSON with source dates, sample counts, and scoring context.
Current endpoints and developer examples.
Country files, historical archives, formats, and source licenses.
Bring the observatory into compatible AI clients.
Read the participation and privacy details before measuring.
Engineering archive
Technical notes on measurement, classification, and distribution. These articles preserve earlier designs and experiments; their historical metrics and architecture descriptions are not current service guarantees. Use the linked source records above for present coverage and the current methodology for model limitations.
Reuse & citation
Voidly-original scores, incidents, and annotations use CC BY 4.0. Raw upstream material retains its own terms, including CC BY-NC-SA 4.0 for OONI raw data and Citizen Lab test lists. Software has separate licenses.
Replace the access date with the day you retrieved the data. Include the source record and observation date for a specific finding.
AI Analytics. (2026). Voidly public censorship index [Dataset]. https://api.voidly.ai/data/censorship-index.json (CC BY 4.0). Accessed YYYY-MM-DD.
@dataset{voidly_2026,
author = {{AI Analytics}},
title = {Voidly public censorship index},
year = {2026},
url = {https://api.voidly.ai/data/censorship-index.json},
note = {Accessed YYYY-MM-DD},
license = {CC BY 4.0}
}