Voidly

Incident record · China

Great Firewall baseline

Type
censorship
Status
confirmed
Severity
critical

Source summary

Anomaly rate: 100.0% (22 measurements).

Open incident JSON
Reported anomaly rate
100%

Source-defined signal; not a universal test-failure percentage.

Source measurement count
22

The record’s count can represent aggregate inputs or alerts.

Incident confidence
70.6%

A source score, not certainty of intentional censorship.

Linked evidence has its own count and confidence fields. These quantities are not interchangeable.

Follow the evidence

Evidence unavailable.

Inspect the observation, the upstream claim and why it was linked.

Qualifying linked evidence sources: Unavailable. Historical incident source labels may include quarantined observations and are not detection evidence by themselves.

The incident response did not include an evidence array. No evidence count is inferred from that absence.

Upstream claims are attributed to their source. Multiple records or source names do not by themselves establish independent corroboration.

Source-reported corroboration

The source’s corroboration assessment is separate from incident confidence and evidence-link confidence. Inspect its source list, count and description.

No corroboration assessment was supplied.

Declared incident sources (historical metadata): OONI · Voidly ML.

Qualifying linked evidence sources: Unavailable.

Fewer than two qualifying linked source labels are present. Stored incident labels may include quarantined OONI observations and do not establish corroboration.

Scope and attribution

What the record names.

Methodology

Affected services

Web connectivityHTTPS

Incident mechanism

No mechanism specified at the incident level.

Individual evidence may carry more specific methods. Those do not automatically describe the entire incident.

Affected domains · 0

Domains named by the source record. Each link opens country-specific OONI measurements; a named domain is not a blanket national-block verdict.

No domains named.

Networks and evidence attribution

ASNs named by this source record: Unavailable.

This aggregate describes fields literally present in qualifying linked evidence. It does not identify who caused an event or prove intent.

No evidence-attribution aggregate was supplied.

What this record cannot establish

Measurements alone do not attribute an event to a government, ISP or another actor, and do not establish intent. An outage or anomalous result can have several explanations.

The country label is an aggregation scope. It is not a census of every network, city or service. Regional or ASN fields belong to their individual source evidence and should not be expanded into a national claim.

A confirmed duration is not supplied. The source observation window and evidence timestamps do not prove an exact start and stop of blocking.

Signal definitions, baseline rates and model versions depend on the source. This record does not establish a common baseline or calibrated certainty. Use the source’s specific signal definition and evaluation before interpreting a score.

Inspect measurement coverage

Cite this incident

Take the record
with you.

Use the permanent published URL.
The original source title is preserved.

The record over time.

Created
Date unavailable
Last updated
Date unavailable
Revision
Unavailable
Revision history and original fields

Revision-history array unavailable.

Canonical ID: 396da7851211. Source ID: 396da7851211. Hash ID: Unavailable.

Additional source duration field: unavailable. It is retained separately from the minutes field and does not establish a verified event duration.

Complete source JSON
{
  "affectedDomains": [],
  "affectedServices": [
    "Web connectivity",
    "HTTPS"
  ],
  "anomalyRate": 1,
  "confidence": 0.7054999999999999,
  "country": "CN",
  "countryName": "China",
  "description": "Anomaly rate: 100.0% (22 measurements).",
  "endTime": null,
  "flag": "🇨🇳",
  "id": "396da7851211",
  "incidentType": "censorship",
  "measurementCount": 22,
  "mechanism": null,
  "severity": "critical",
  "sources": [
    "OONI",
    "Voidly ML"
  ],
  "startTime": "2026-09-03T00:00:00Z",
  "status": "confirmed",
  "title": "Great Firewall baseline"
}
All China incidents

Related records come from up to 100 recent records in the same country. They do not establish a causal connection.

Check this live

Read incident JSON

Read the published record and its observation dates. A fresh response does not mean a new measurement or confirm a current restriction.

MCP and curl examples

Claim assessment is on hold. While the public incident projection is under review, the unified MCP service may answer verify_claim with an insufficient-evidence response. The example below is a request, not a verified conclusion.

Unified MCP endpoint: https://api.voidly.ai/mcp. Use the request in an initialized MCP client. Nothing runs on this page.

The curl command reads the published incident record; it does not call verify_claim or test a network connection.

MCP endpoints and availability

Agent summary

Facts: Great Firewall baseline. Country: China. Source type: censorship. Source status: confirmed. Reported observation window: 2026-09-03T00:00:00Z to unavailable.

As of linked evidence: Unavailable. Record generated: unavailable; page checked: 2026-11-22T22:57:47.953Z. These are different times.

Stored incident source labels: 2 distinct declared names (OONI, Voidly ML). These are historical metadata; qualifying linked sources are shown above. Source names are not independent corroboration.

Cite: Voidly Research. (n.d.). Great Firewall baseline [396da7851211]. Voidly Global Censorship Index. https://voidly.ai/incident/396da7851211

Publication date: Not supplied.

Use this in your agent

Connect the read-only Atlas MCP service, then select and copy this request into your agent.

Atlas MCP setup · Agent summary JSON · Endpoint: https://atlas-mcp.voidly.ai/mcp

Source generated:
Page source check:

Permanent incident URL

Voidly-original annotations: CC BY 4.0. Upstream data retain their source licenses.