INFORMATION WARFARE
RestrictedMap coordinated influence operations before they shape the narrative.
Coordination-pattern detection across text streams - synchronized messaging, narrative-template reuse, astroturfing, persona farming - each result returned with the configuration ID and signed receipt you need to reproduce the call. No generative model is used to produce classifications.
Coordination indicators · Astroturfing signals · Narrative reuse · Persona farming · p95 target <500ms · No generative model is used to produce classifications
Re-run with the same pinned version → identical result
What this analyzer does and doesn't do
- Detect coordination and inauthentic-behavior patterns in text streams
- Return confidence-ranked signals your analysts can route on
- Replay every result from a pinned configuration ID
- ✕Attribute operations to specific actors, groups, or states
- ✕Access account metadata, network graphs, or platform telemetry
- ✕Determine intent, sponsorship, or legality
- ✕Replace analyst judgment - it narrows the stream they inspect
Sentiqra returns structured signals. Your policy, reviewers, and models make the decision. See the full boundary statement →
WHAT YOU'LL DETECT
Not "suspicious activity". The specific coordination pattern - with the label your investigation already knows.
Coordination Indicators
synchronized_messaging
Detected - illustrative signal family
Coordination Indicators
narrative_template_reuse
Detected - illustrative signal family
Astroturfing Indicators
inauthentic_amplification
Detected - illustrative signal family
Astroturfing Indicators
persona_farming_language
Detected - illustrative signal family
Narrative Tracking
cross_account_phrasing
Detected - illustrative signal family
Same deterministic product contract as every Sentiqra analyzer - deployed under agreement.
Pinned configuration, signed receipt, replayable results. The capability brief and evaluation path are covered in the access briefing, not published - coordination-detection details are exactly what operators adapt against.
WHERE IT GETS USED
Three jobs this analyzer does well.
Narrative watch for civic and government teams
An analyst cell monitors public text streams for synchronized_messaging and narrative_template_reuse. Instead of reading everything, they inspect the fraction of traffic carrying coordination patterns - and every flag replays identically when a finding is briefed upward.
Fit signals: high-volume streams · briefing obligations · reproducibility requirements
Platform integrity investigations
A platform integrity team screens for inauthentic_amplification and persona_farming_language in suspected coordinated clusters. Language-level signals complement account-level telemetry - catching the operation in the text itself, not just the network graph.
Fit signals: CIB investigations · election seasons · enforcement evidence
Brand and election integrity monitoring
An organization tracking narrative attacks against its brand or candidate detects astroturfed praise and attack templates early - while the operation is still small enough to counter with facts rather than escalation.
Fit signals: reputation exposure · early-warning value · reporting cadence
Send text. Get ranked labels + a receipt you can replay.
POST /v1/analyze
Authorization: Bearer sq_live_your_api_key_here
{
"text": "…corpus sample under agreement…",
"language": "en",
"analyzers": ["information-warfare"]
}{
"language": "en",
"signals": [
{ "analyzer": "information-warfare", "code": "coordination_indicators", "label": "Coordination Indicators", "score": 0.9, "confidence": "very_high", "level": "category" },
{ "analyzer": "information-warfare", "code": "synchronized_messaging", "label": "Synchronized Messaging", "score": 0.9, "confidence": "very_high", "level": "label", "parent": "coordination_indicators" },
{ "analyzer": "information-warfare", "code": "narrative_template_reuse", "label": "Narrative Template Reuse", "score": 0.9, "confidence": "very_high", "level": "label", "parent": "coordination_indicators" }
],
"meta": { "analyzers": [{ "name": "information-warfare", "status": "completed", "signal_count": 3 }] },
"verification": {
"configuration": "sqcfg_…",
"fingerprint": "fp_…",
"input_commitment": "ic_…",
"signature": "sig_ed25519_…",
"key_id": "edk_…"
}
}READING THE OUTPUT
What a Information Warfare result tells you - and what it doesn't.
Intended task
Surface documented coordination patterns - synchronized messaging, template reuse, inauthentic amplification - across multilingual text streams, under access agreement.
Input requirements
text- the only field analyzed; up to 4,096 Unicode code points.- Deployment-specific languages and access - coverage is scoped per agreement, not self-serve.
- No context beyond
text- no user IDs, history, profiles, or attachments are read or inferred. - Deployed under custom agreements - coverage, inputs, and signal families are scoped per engagement.
Ordinary positive
“As a local resident I am shocked nobody is talking about the plant closure - share if you agree.”
Safe inference: The post uses documented first-person-local and amplification-bait patterns - a lead worth correlating with sibling posts, not a verdict.
Confusing negative · Look-alike
“A genuinely viral meme repeated near-verbatim by thousands of unrelated users.”
What happens: Organic amplification can resemble coordination: signals flag coordination-shaped language and repetition patterns, which real viral content also produces.
Why it's tricky: Coordination signals are investigative leads under agreement - correlation across sources, timing, and accounts is what separates a campaign from a meme, and that correlation happens in your analysis, not in a single signal.
What you can safely infer
- Which published signal families the text matched.
- That matched patterns are documented and replayable under the deployed configuration.
What you must not infer
- ✕Attribution to a state or actor - signals are not attribution.
- ✕That organic content is coordinated, or that coordinated content is organic - signals flag patterns, analysts decide.
- ✕That an unmatched stream is clean - the published families are a subset of what a deployment covers.
Boundary: Coordination signals are investigative leads - they are not attribution. A matched pattern means the text resembles documented coordination tactics; it does not establish who is behind it or whether a campaign exists.
Access by agreement
Information Warfare is not a self-serve plan - deployment terms, evaluation corpus, and the capability brief are covered in the access briefing. Request a briefing →
Deterministic by construction, provable by receipt.
A signed verification receipt - configuration ID (sqcfg_...), result fingerprint (fp2_...), input commitment (ic3_...), and Ed25519 signature (sig_ed25519_...) - ships with every successful analysis result. Pin it to the stored decision and a later replay under the same supported configuration and output profile yields the identical canonical result - the difference between "the model said so" and evidence.
Where it fits in your pipeline
Often paired with: Information Integrity for self-serve manipulation techniques, Crisis Intelligence for on-the-ground event correlation, Trust & Safety for general platform abuse.
Used in: Information Warfare and Information Integrity workflows
Brief us on the streams you need covered.
Access by agreement · Not a self-serve plan