🔮 Predictive Military Intelligence 30/60/90-day estimate bands · low / mid / high · confidence + methodology disclosed ← Back to MIL Intel
▸ FORECAST PHILOSOPHY · METHODOLOGY DISCLOSURE
Every row is a triplet low / mid / high around an explicit base — the "current_confirmed" pip. Confidence is the analyst-assigned likelihood that the realised value falls inside the band: HIGH = tight band, MED = moderate, LOW = wide uncertainty. Methodology is named per row (linear_extrapolation, expert_estimate, scenario_analysis, …) and every estimate carries citations in evidence_json. We disclose this so consumers do not treat the mid as a point prediction.
Latest forecast run: 2026-10-28 · 8 categories
🛸 Drones
350
[298–438] HIGH
🛡 Air defense
100
[85–125] HIGH
⚓ Naval
15
[13–19] HIGH
👤 Personnel
343
[292–429] LOW
🌐 Foreign advisors
1
[1–1] LOW
📡 SIGINT
7
[6–9] HIGH
💻 Cyber
106
[90–132] LOW
🎯 Threat probability
38
[32–48] LOW
🛸
Drones / Drones
HIGH
count
L: 298 M: 350 H: 438
→ stable Δ30d: 0 conf: 350 [linear_extrapolation+public_attribution]
Linear extrapolation from military_inventory DRONE rows (350 confirmed) + 0 pattern_alert drone mentions 90d; CSIS / OE Watch attribution of Iranian/Russian deliveries.
30D TREND (estimate_mid)
▸ evidence (3)
  • ↗ 2026-03-15 — Cuba drone build-up — Iranian Mohajer-6 and Shahed-loitering munition transfers via Russian shipping documented 2024-2026.
  • ↗ 2026-02-01 — Cuban FAR drone inventory expansion — public training-school imagery, multiple squadrons stood up.
  • ↗ 2026-10-28 — military_inventory DRONE confirmed-sum=350; pattern_alerts drone mentions 90d=0; vessel_movements drone-cargo 180d=0
🛡
Air defense / Defensa aérea
HIGH
capability_score
L: 85 M: 100 H: 125
→ stable Δ30d: 0 conf: 100 [public_attribution+linear_extrapolation]
Composite capability score (SAM×2 + radar×5 + RU/CN/IR×3, capped 100). IISS Mil Balance + CSIS modernization attribution.
30D TREND (estimate_mid)
▸ evidence (3)
  • ↗ 2026-01-01 — Cuba SAM/radar holdings: S-125 Neva/Pechora SA-3, S-75 Volga SA-2 modernization, P-18/P-19 radars — Russian-pattern supplier.
  • ↗ 2026-04-10 — Air-defense modernization signs — Russian technician rotations and PLAAF advisory visits 2025-2026.
  • ↗ 2026-10-28 — SAM/missile sum=1180; radar rows=1; Russian/Chinese/Iranian rows=0
⚓
Naval / Naval
HIGH
active_hulls
L: 13 M: 15 H: 19
→ stable Δ30d: 0 conf: 15 [shipping_inference]
Count of MILITARY-category Cuban hulls in vessels table (15); 0 distinct hulls observed in vessel_movements last 30d.
30D TREND (estimate_mid)
▸ evidence (3)
  • ↗ 2026-01-01 — Cuban Revolutionary Navy (MGR): ~13 patrol craft / coast-guard / SIGINT hulls, primarily Soviet-era Pauk II + Stenka.
  • ↗ 2025-12-05 — MGR SIGINT vessels CAMILO CIENFUEGOS and VILLA CLARA monitored on GTMO approaches.
  • ↗ 2026-10-28 — Cuban-flagged MILITARY hulls=15; observed in vessel_movements 30d=0
👤
Personnel / Personal
LOW
active_count
L: 292 M: 343 H: 429
→ stable Δ30d: 0 conf: 343 [linear_extrapolation+analyst_consensus]
Persons table rank-keyword match yields 98 military-rank entries; reservist factor ×2.5 (IISS Mil Balance historical) = 245 reservists; combined active+reserve = 343.
30D TREND (estimate_mid)
▸ evidence (2)
  • ↗ 2026-01-01 — Cuba active-duty estimate ~50,000 + reserve ~39,000; Servicio Militar General mandatory conscription.
  • ↗ 2026-10-28 — Military-rank persons rows=98; reservist factor 2.5 applied (per IISS).
🌐
Foreign advisors / Asesores extranjeros
LOW
count
L: 1 M: 1 H: 1
→ stable Δ30d: 0 conf: 0 [shipping_inference+public_attribution]
Sum of foreign mil-air movements 90d (0) + foreign-flag mil vessel movements 90d (0) + advisor-tagged alerts 90d (0). Atlantic Council + CSIS PLA/GRU presence attribution.
30D TREND (estimate_mid)
▸ evidence (3)
  • ↗ 2026-04-22 — Russian, PRC, Iranian and Venezuelan advisory rotations to Cuba — port calls and chartered air shuttles 2024-2026.
  • ↗ 2026-03-15 — PLA / GRU presence at Bejucal, Wajay, Lourdes successor SIGINT facilities — open-source corroboration.
  • ↗ 2026-10-28 — Foreign aircraft mvmts 90d=0; foreign-flag mil vessel mvmts 90d=0; advisor-tagged alerts 90d=0
📡
SIGINT / SIGINT
HIGH
sites_active
L: 6 M: 7 H: 9
→ stable Δ30d: 0 conf: 7 [public_attribution]
strategic_sites rows with category in (SIGINT/SPY_BASE/ELINT) = 7. CSIS + WSJ open-source confirmation of Bejucal/El Salao/Wajay capacity expansion.
30D TREND (estimate_mid)
▸ evidence (3)
  • ↗ 2024-07-01 — Bejucal / El Salao / Wajay / Calabazar / Paseo signals-intelligence facilities — capacity expansion 2023-2026.
  • ↗ 2023-06-08 — PRC eavesdropping facilities in Cuba — confirmed by US officials, expansion underway.
  • ↗ 2026-10-28 — strategic_sites SIGINT/SPY_BASE/ELINT active rows=7
💻
Cyber / Ciber
LOW
active_operatives
L: 90 M: 106 H: 132
→ stable Δ30d: 0 conf: 106 [linear_extrapolation+analyst_consensus]
regime_developers tier=HIGH count 38 + iceberg factor ×1.8 (known ratio of public-attributable to suspected total) = 106. OFAC press + internal DB attribution.
30D TREND (estimate_mid)
▸ evidence (3)
  • ↗ 2026-05-07 — Sanctions enforcement against Cuban cyber operatives tied to UCI / XETID — designations 2024-2026.
  • ↗ 2026-05-18 — Codersrank Cuba leaderboard + UCI / XETID / Etecsa repo footprint — tier=HIGH operatives confirmed in DB.
  • ↗ 2026-10-28 — regime_developers tier=HIGH rows=38; iceberg factor ×1.8
🎯
Threat probability / Probabilidad de amenaza
LOW
P_30d_pct
L: 32 M: 38 H: 48
→ stable Δ30d: 0 conf: 38 [analyst_consensus+linear_extrapolation]
Aggregate 30-day fire probability across the 13 tracked threat patterns; total fires last 90d = 350, mean P_30d = 38%.
30D TREND (estimate_mid)
▸ evidence (2)
  • ↗ 2026-05-18 — Historical 90-day fire-rate of the 13 tracked threat patterns — derived from pattern_alerts.last_seen and occurrence_count.
  • ↗ 2026-10-28 — 13-pattern aggregate fires last 90d=350; mean P_30d=38%
← Back to MIL Intel · methodology: model-based projection · archive: query mil_forecasts by forecast_date