Scrutica
Cross-program decomposition
A 6-pattern taxonomy applied to each program’s announced-vs-deployed gap. Each program’s gap decomposes into one or more named patterns whose percentages sum to 100% of the program’s gap. The taxonomy itself is a Scrutica editorial framework; per-row attributions carry source URL + accessed-at date + confidence rating. The per-program inflation_decomposition already shipped on the per-country cards (financing-type breakdown: government / private / FDI / credit) is a separate decomposition — this scorecard surfaces the gap-shape patterns that financing-type breakdown does not capture.
A cross-program reading of whyannounced sovereign-AI numbers outrun deployed reality. Each program’s announced-vs-deployed gap falls into one or more named patterns: Aspirational (the announcement is a future-target capacity), Phased rollout (Phase 1 is real, Phase N is forward-scheduled), Financing-stalled (intent is real, capital closure isn’t), Chip-allocation-conditional (deployment is authorization-gated), Vaporware (no substantive substrate), Other. The framework is a Scrutica editorial choice; readers who would classify a row differently can read the per-row notes inline.
Programs
34
34 sovereign vehicles tracked
With announced-vs-deployed gap
28
6 executing without material gap
Primary-source attributions
28
of 62 total — confidence tier 1
Editorial attributions
5
Scrutica-editorial (tier 3) where substrate is sparse
Portfolio frequency
Aspirational
16/ 28
7 dominant · avg 41% when present
Phased rollout
21/ 28
14 dominant · avg 60% when present
Financing-stalled
9/ 28
0 dominant · avg 26% when present
Per-program decomposition
Each row sums to 100% across the patterns present for that program. Hover or focus a cell to read the per-attribution provenance + notes. Programs without a material gap (committed ≈ disbursed ≈ announced) are listed separately below as “executing”.
Portfolio reading
Phased rollout dominates
14 programs: Australia · Canada · Germany · Spain · EU-Wide · United Kingdom · +8 more
Aspirational dominates
7 programs: UAE · Brazil · France · South Korea · Malaysia · Saudi Arabia · +1 more
Other dominates
5 programs: Africa (Regional) · NATO/AUKUS · Poland · Sweden · Ukraine
Vaporware dominates
1 program: Indonesia
Regional reading. Gulf programs (UAE, KSA) skew Aspirational + Chip-allocation-conditional — headline 5 GW / 600K-GPU framing with operationally smaller Phase 1 + a BIS-authorization gate post-Oct 2023. South & Southeast Asia (India, Indonesia, Vietnam, Malaysia) skew Phased (where execution is real) versus Financing-stalled or Vaporware (where capital + counterparties aren’t yet identified). Western Europe (France, Germany, UK, EuroHPC) skews Phased with smaller Aspirational residuals — the €109B France headline is the outlier that drags the regional average toward aspirational. East Asia (China, Japan, Korea) skews Chip- allocation-conditional (China specifically) and Phased (Japan, Korea).
Framework + methodology
The 6-pattern taxonomy is a Scrutica editorial framework, not an industry standard. Categories are not mutually exclusive. Per-pattern percentages partition the announced-versus-deployed gap (so attributions for a single program sum to 100% of the gap, with the gap itself implicit in committed/disbursed substrate). Where a program has no gap (disbursed ≈ announced), the row shows pct=0 across patterns and a no_gap flag. Where the literature is silent, the program is marked unclassified with an explicit reason.
Aspirational
Public announcement of a future-target capacity with no operational deployment yet, OR the announced number includes a capacity target whose construction has not commenced. Phase 1 may be commissioned, but the headline-shaping bulk is forward-target.
Example: UAE Stargate $500B (5 GW campus aspiration; only Phase 1 200MW under construction)
Phased rollout
Multi-phase plan where Phase 1 (or earlier) is operational and on schedule; Phase N is forward-target but not aspirational because the multi-year build cadence is itself the disbursement plan. Distinguished from Aspirational by the presence of a substantiated multi-phase budget schedule.
Example: IndiaAI Mission (38K GPUs operational, 100K target end-2026, on track per IndiaAI CEO disclosure)
Financing-stalled
Announcement carries deployment intent but financing closure has not arrived — SWF allocation pending, sovereign-debt instrument not yet placed, parliamentary appropriation deferred, chip-supplier contract un-priced. Distinguished from Aspirational by the presence of operational intent rather than future-capacity branding.
Five trackers / publications that overlap the sovereign-AI lens. Scrutica — Sovereign AI Reality Dashboard tracks 34 programs with per-program inflation decomposition, interdependence scoring, and disclosure-state reconciliation; competitor cells label as verified-absent / verified-inaccessible / subscription-only per anti-fabrication discipline.
CNAS — Technology & National Security publishes ad-hoc reports rather than a continuous country-level tracker — the comparison establishes coverage gaps, not methodology deltas.
Access note · No continuous sovereign-AI tracker found at the program page on 2026-05-20. CNAS publishes ad-hoc reports (Atomic Advantage, Promethean Rivalry, Digital Silk Road: Brazil) under its TNS sub-programs (America’s AI Leadership, Biopower, Quantum Leap, Countering the Digital Silk Road), but no continuous country-level tracker.
| Dimension | Scrutica | CNAS |
|---|---|---|
| Programs tracked (country-level roster) | 34programs · per-country “reality ratio” (disbursed / announced government-only). | Ad-hoc reports per country/region; no continuous roster.Not published sourceaccessed 2026-05-20 |
| Inflation decomposition (announced vs deployed) | Per-program component breakdown: government-only spending vs FDI / private commitment / credit-line padding. | Not published.Not published sourceaccessed 2026-05-20 |
CNAS — Technology & National Security publishes ad-hoc reports rather than a continuous country-level tracker — the comparison establishes coverage gaps, not methodology deltas.
Switch to Analyst view (top of page) for the side-by-side methodology table.
Landing page not directly accessible at fetch time (2026-05-20). Cells reflect Scrutica's research scan; correct upward if a continuous tracker is identified.
Access note · CSIS topic landing page returned ECONNREFUSED at fetch 2026-05-20; Wadhwani AI Center publishes individual reports on AI and national strategy but Scrutica’s scan found no continuous country-level sovereign-AI tracker. Cell states reflect that finding; correct upward if a tracker is identified.
| Dimension | Scrutica | CSIS |
|---|---|---|
| Programs tracked (country-level roster) | 34programs · per-country “reality ratio” (disbursed / announced government-only). | Topic-specific reports; no continuous roster identified. |
Brookings — AI Topic publishes ad-hoc reports rather than a continuous country-level tracker — the comparison establishes coverage gaps, not methodology deltas.
Access note · Brookings’ AI topic page surfaces recent commentary (May 2026) but Scrutica’s scan found no dedicated sovereign-AI tracker, multi-country index, or comparative dataset on 2026-05-20.
| Dimension | Scrutica | Brookings |
|---|---|---|
| Programs tracked (country-level roster) | 34programs · per-country “reality ratio” (disbursed / announced government-only). | Commentary + Center for Technology Innovation policy pieces; no continuous tracker.Not published |
Subscription-only — Scrutica cannot row-by-row verify IDC — Sovereign AI market sizing's coverage from public materials; the cells below mark what is publicly observable.
Access note · IDC Sovereign AI market-sizing research is sold via subscription and is not directly accessed by Scrutica. The /sovereign-ai inclusion threshold (≥ $100M committed OR named in CRS / NVIDIA earnings / IDC sovereign-AI tracking) treats IDC as an authority on whether a program is real, not as a published dataset Scrutica can compare row-by-row.
| Dimension | Scrutica | IDC |
|---|---|---|
| Programs tracked (country-level roster) | 34programs · per-country “reality ratio” (disbursed / announced government-only). |
Epoch AI — Hyperscaler concentration occupies the closest adjacent lens; the row-by-row delta below shows where Scrutica extends the company-level concentration view into per-country sovereign-program tracking.
Access note · Epoch publishes per-company compute concentration (Amazon, Google, Meta, Microsoft, Oracle — 71% of cumulative AI compute as of Q4 2025), not per-country sovereign-AI program tracking. The lens is complementary, not equivalent.
| Dimension | Scrutica | Epoch AI |
|---|---|---|
| Programs tracked (country-level roster) | 34programs · per-country “reality ratio” (disbursed / announced government-only). | Five hyperscalers + named chip manufacturers. NOT country-level. |
Chip-allocation-conditional
2/ 28
1 dominant · avg 30% when present
Vaporware
1/ 28
1 dominant · avg 50% when present
Other
7/ 28
5 dominant · avg 79% when present
Example: Indonesia Sovereign AI Fund planned 2027-2029, not established
Chip-allocation-conditional
Announcement contingent on BIS / chip-supplier approval (export license, NVIDIA allocation slot, foundry-capacity commitment). Distinguished from financing-stalled by the constraint being a supplier or regulatory authorization rather than capital.
Example: UAE G42 deployment contingent on continued US chip export approval (received Nov 2025); subsequent ramps remain authorization-gated
Vaporware
Announcement with no substantive substrate — no facility, no operator, no allocation. Distinguished from financing-stalled by the absence of identifiable counterparties.
Example: Various smaller-state announcements where the program-name does not resolve to a procurement vehicle
Other
Catch-all for cases that don't fit (joint-venture-pending, regulatory-blocked, dual-currency-conversion ambiguity, sovereign-political-instability hold). Notes inline.
Example: Program where deployment intent is real, financing exists, but a downstream condition unrelated to the five named patterns is the binding constraint
Cross-program sovereign-AI inflation framing exists at CNAS (Sovereign AI Index, periodically updated), Brookings (sovereign-AI capacity assessments), CSIS (sovereign technology programs), and IDC (sovereign AI tracker). None of these works publish a per-program decomposition along Scrutica’s 6-pattern taxonomy — closest analogues are CNAS’s “announced versus disbursed” pairs and IDC’s “announced versus operational” pairs, which both surface the gap as a single ratio without naming the gap’s shape. The Scrutica framework is a Scrutica editorial choice positioned as such; a reader who would prefer a different taxonomy can read the per-row notes and re-classify inline. Source citations per attribution accompany every row in the scorecard above.
| Interdependence matrix (NVIDIA / TSMC / US / BIS reach) |
Composite governance-reach score derived from chip-dependency × cloud-dependency × BIS-reach × non-aligned-hop. |
Not published as a continuous matrix.Not published sourceaccessed 2026-05-20 |
| Provenance per program (authority_tier per claim) | Inline source list + authority_tier (1 = government disclosure; 3 = press) per program record. | Per-report citations; not normalized across countries.Not published sourceaccessed 2026-05-20 |
| Update cadence | Substrate refresh on each ingest cycle (May 2026 tier-2 sovereign ingest added HUMAIN, Yotta, EuroHPC, Tencent, Baidu, Huawei, ByteDance, JD, Kingsoft, China Telecom). | Ad-hoc / per-report.Not published sourceaccessed 2026-05-20 |
| Open data export | CSV / JSON page-level export; CC-BY 4.0 (matches Epoch + CSET licensing). | PDF reports; no machine-readable dataset.Not published sourceaccessed 2026-05-20 |
| Inflation decomposition (announced vs deployed) | Per-program component breakdown: government-only spending vs FDI / private commitment / credit-line padding. | Not identified.Inaccessible sourceaccessed 2026-05-20 |
| Interdependence matrix (NVIDIA / TSMC / US / BIS reach) | Composite governance-reach score derived from chip-dependency × cloud-dependency × BIS-reach × non-aligned-hop. | Not identified.Inaccessible sourceaccessed 2026-05-20 |
| Provenance per program (authority_tier per claim) | Inline source list + authority_tier (1 = government disclosure; 3 = press) per program record. | Not identified.Inaccessible sourceaccessed 2026-05-20 |
| Update cadence | Substrate refresh on each ingest cycle (May 2026 tier-2 sovereign ingest added HUMAIN, Yotta, EuroHPC, Tencent, Baidu, Huawei, ByteDance, JD, Kingsoft, China Telecom). | Ad-hoc / per-report.Inaccessible sourceaccessed 2026-05-20 |
| Open data export | CSV / JSON page-level export; CC-BY 4.0 (matches Epoch + CSET licensing). | PDF reports; no machine-readable dataset identified.Inaccessible sourceaccessed 2026-05-20 |
Landing page not directly accessible at fetch time (2026-05-20). Cells reflect Scrutica's research scan; correct upward if a continuous tracker is identified.
Switch to Analyst view (top of page) for the side-by-side methodology table.
| Inflation decomposition (announced vs deployed) | Per-program component breakdown: government-only spending vs FDI / private commitment / credit-line padding. | Not published.Not published sourceaccessed 2026-05-20 |
| Interdependence matrix (NVIDIA / TSMC / US / BIS reach) | Composite governance-reach score derived from chip-dependency × cloud-dependency × BIS-reach × non-aligned-hop. | Not published.Not published sourceaccessed 2026-05-20 |
| Provenance per program (authority_tier per claim) | Inline source list + authority_tier (1 = government disclosure; 3 = press) per program record. | Per-piece citations; not normalized.Not published sourceaccessed 2026-05-20 |
| Update cadence | Substrate refresh on each ingest cycle (May 2026 tier-2 sovereign ingest added HUMAIN, Yotta, EuroHPC, Tencent, Baidu, Huawei, ByteDance, JD, Kingsoft, China Telecom). | Continuous commentary; no scheduled tracker refresh.Not published sourceaccessed 2026-05-20 |
| Open data export | CSV / JSON page-level export; CC-BY 4.0 (matches Epoch + CSET licensing). | PDF reports + blog posts; no dataset.Not published sourceaccessed 2026-05-20 |
Brookings — AI Topic publishes ad-hoc reports rather than a continuous country-level tracker — the comparison establishes coverage gaps, not methodology deltas.
Switch to Analyst view (top of page) for the side-by-side methodology table.
| Inflation decomposition (announced vs deployed) | Per-program component breakdown: government-only spending vs FDI / private commitment / credit-line padding. | Not directly verifiable from public materials.Subscription-only sourceaccessed 2026-05-20 |
| Interdependence matrix (NVIDIA / TSMC / US / BIS reach) | Composite governance-reach score derived from chip-dependency × cloud-dependency × BIS-reach × non-aligned-hop. | Not directly verifiable.Subscription-only sourceaccessed 2026-05-20 |
| Provenance per program (authority_tier per claim) | Inline source list + authority_tier (1 = government disclosure; 3 = press) per program record. | Not directly verifiable.Subscription-only sourceaccessed 2026-05-20 |
| Update cadence | Substrate refresh on each ingest cycle (May 2026 tier-2 sovereign ingest added HUMAIN, Yotta, EuroHPC, Tencent, Baidu, Huawei, ByteDance, JD, Kingsoft, China Telecom). | IDC research cadence (quarterly market updates).Subscription-only sourceaccessed 2026-05-20 |
| Open data export | CSV / JSON page-level export; CC-BY 4.0 (matches Epoch + CSET licensing). | Subscription-only PDFs; no open export.Subscription-only sourceaccessed 2026-05-20 |
Subscription-only — Scrutica cannot row-by-row verify IDC — Sovereign AI market sizing's coverage from public materials; the cells below mark what is publicly observable.
Switch to Analyst view (top of page) for the side-by-side methodology table.
| Inflation decomposition (announced vs deployed) | Per-program component breakdown: government-only spending vs FDI / private commitment / credit-line padding. | Not applicable (company-level lens; no announced-vs-deployed concept).Not published sourceaccessed 2026-05-20 |
| Interdependence matrix (NVIDIA / TSMC / US / BIS reach) | Composite governance-reach score derived from chip-dependency × cloud-dependency × BIS-reach × non-aligned-hop. | Per-vendor chip-source matrix (Nvidia / Google / AMD / Huawei / Amazon). sourceaccessed 2026-05-20 |
| Provenance per program (authority_tier per claim) | Inline source list + authority_tier (1 = government disclosure; 3 = press) per program record. | AI Chip Owners explorer underlies the concentration figure (CC-BY 4.0). sourceaccessed 2026-05-20 |
| Update cadence | Substrate refresh on each ingest cycle (May 2026 tier-2 sovereign ingest added HUMAIN, Yotta, EuroHPC, Tencent, Baidu, Huawei, ByteDance, JD, Kingsoft, China Telecom). | Quarterly updates to chip-owners dataset. sourceaccessed 2026-05-20 |
| Open data export | CSV / JSON page-level export; CC-BY 4.0 (matches Epoch + CSET licensing). | CC-BY 4.0; CSV download per record. sourceaccessed 2026-05-20 |
Epoch AI — Hyperscaler concentration occupies the closest adjacent lens; the row-by-row delta below shows where Scrutica extends the company-level concentration view into per-country sovereign-program tracking.
Switch to Analyst view (top of page) for the side-by-side methodology table.