Azimuth.reportSpecial ReportAZ-SR-2026-10-SWARM
Swarm-AI Attack Register
Kinetic, cyber and influence operations in which coordinated AI agents or unmanned systems were used, claimed, or assessed — each graded for source quality and swarm evidence.
Track
AI-swarm status
Timeline by start of activity · mark size = Evidence Strength Index
Register
| ID | Start | Track | Incident | AI-swarm status | Grade | ESI | Autonomy |
|---|
Status ladder
Confirmed requires first-hand disclosure plus independent corroboration of the AI-coordinated element. Probable rests on one first-hand technical source. Possible, Claimed and Unverified step down from there. Excluded entries are frequently called swarms but assessed otherwise.
Grading
Admiralty code (source A–F, information 1–6) per entry, plus ICD-203 confidence. Event status is graded separately from the AI-swarm claim: an attack can be certain while its AI element is not.