← Back to Censorship Index

Methodology

How we measure global internet censorship

Version 2.0Updated: 2026-02-19JSON

Overview

Composite score from three data sources, processed through federated ML pipeline, updates continuously based on live network measurements.

When a government blocks a new service, our data reflects the change within hours.

Data Sources

OONI Measurements

Samples14,204,087
Coverage50 countries
TestsWeb, Messaging, Circumvention

Sensor Network

Nodes16
Coverage6 continents
Interval30 seconds

User Telemetry

Users...+
RetentionAggregated only
PrivacyNo individual tracking

Multi-Source Correlation

No single measurement network captures the full picture of internet censorship. OONI provides active probing but has geographic gaps. CensoredPlanet provides remote measurement but lacks ground truth. IODA detects outages but not selective blocking. Cloudflare Radar sees traffic shifts but not which domains are blocked.

Voidly operates its own 16-node probe network across 6 continents — testing VPN accessibility and censorship patterns every 5 minutes — then correlates these proprietary measurements with all four external sources to produce verified incidents with evidence chains. This turns ambiguous network anomalies into structured, citable censorship intelligence.

ML Model

Gradient boosting classifier trained on 37K labeled censorship incidents. Federated learning across 16 nodes ensures no raw user data leaves local systems.

Censorship Classifier

AlgorithmGradientBoosting
F1 Score99.8%
AUC-ROC1.000
Training Samples37K labeled incidents
ScheduleDaily @ 02:00 UTC

Shutdown Forecast Model

AlgorithmXGBoost
ROC AUC74.6%
Recall50%
Training Samples14.6K historical records
ScheduleWeekly (Sundays @ 02:00 UTC)

Feature Importance

country_censorship_score
28%
destination_blocked
22%
node_success_rate
18%
user_country_encoded
12%
node_load
8%
hour_of_day
6%
day_of_week
4%
is_peak_hours
2%

Scoring System

0-100 scale. 0 = complete freedom. 100 = total censorship.

0-10
Free
Minimal or no censorship
11-25
Low
Limited content restrictions
26-45
Medium
Significant restrictions on some platforms
46-70
High
Widespread blocking of platforms and news
71-100
Severe
Pervasive censorship / isolated internet

Limitations

  • Scores are national averages — regional variations not captured
  • VPN detection underreported in highly restricted environments
  • Sample sizes vary by country — affects confidence levels
  • Real-time events may take up to 24h to reflect in scores
  • Content filtering and throttling harder to detect than blocking
  • Self-censorship and legal restrictions not measured

Confidence Intervals

Each country score includes a confidence interval reflecting measurement certainty. Wider intervals indicate less data or greater variability.

Country
Score
Interval
Confidence
Note
China
66%
± 2%
high
Large sample
Iran
42%
± 4%
high
Russia
31%
± 3%
high
Myanmar
21%
± 7%
medium
Smaller sample

Validation

Scores are validated against external benchmarks and known censorship events. Continuous evaluation ensures model accuracy over time.

BaselineFreedom House — Freedom on the Net
Correlationr = 0.87
Ground TruthKnown events (e.g. Iran shutdowns match score spikes)
Cross Validation5-fold
Classifier F199.8%
Classifier AUC-ROC1.000
Forecast AUC-ROC74.6%

Update Pipeline

OONIIngestionFeature EngineeringML ScoringIndex Update
TrainingDaily @ 02:00 UTC
PublicationDaily @ 03:00 UTC
Score Latency~24h
Raw Ingestion~5min

Citation

Use this data in research? Please cite:

APA Format

Voidly Research. (2026). Global Censorship Index. https://voidly.ai/censorship-index

BibTeX

@misc{voidly_censorship_index,
  author = {Voidly Research},
  title = {Global Censorship Index},
  year = {2026},
  url = {https://voidly.ai/censorship-index}
}

License: CC BY 4.0 — Free to use with attribution

Data Access

Contact