Evidence for compute governance
AI infrastructure records for research on ownership, supply dependencies and compute policy.
Why Scrutica exists
A facility’s power capacity alone says little about the training it can support. The answer depends on its hardware and utilization; regulatory reach also depends on the companies behind it and the jurisdictions they operate in.
What Scrutica covers
Sources include Epoch AI’s facility and GPU-cluster datasets, the Federal Register, SEC EDGAR, CSET ETO and licensed corporate databases. The methodology page describes their coverage and the gaps that affect the analysis.
What's in the platform
Records cover the physical infrastructure and the organizations responsible for it.
- FacilitiesData centers and fabrication plants across 110 countries, with location, operator, power capacity and GPU deployment where available.
- Supply chain relationships14,937 distinct supplier–customer relationships, counted separately by product or service after duplicate company identities and repeated reports are reconciled. Corporate ownership is recorded separately.
- Sovereign AI programs36 government compute programs, with announced funding, commitments and disbursements recorded separately where disclosed.
- Export controls3,435 BIS Entity List designations, 176 active designation-to-company matches. A designation may match more than one company record.
- Compute capacity estimatesCapacity estimates from hardware inventories, power capacity and capital cost, with adjustable assumptions and uncertainty bounds.
The Query Builder answers questions over these layers in natural language, using 55 tools for records, calculations and page content. The MCP server exposes 10 read-only tools for use in an agent client.
Which number to cite
The platform totals below use the snapshot dated 2026-09-11. Individual pages and Query Builder results may use a different date or population. Cite the result you used, with its date and scope.
| Quantity | Value | Coverage | Vintage | Definition | Current comparison |
|---|---|---|---|---|---|
| Facilities | 4,576 | all recorded facilities, across classes and statuses; includes records from licensed sources that are not publicly readable | 2026-09-11 | the substrate snapshot | Publicly readable facilities: 4,260 at last render. Public access excludes the licensed-source records included in the total. |
| Countries | 110 | distinct country codes in facility records | 2026-09-11 | the substrate snapshot | Counted when the snapshot was created; no current comparison is run here. |
| Supply-chain records | 18,980 | all supply-chain records, including licensed sources and repeated reports of the same relationship | 2026-09-11 | the graph scope | The underlying records have restricted access. This total is counted when the snapshot is created. |
| Distinct supplier–customer relationships | 14,937 | distinct combinations of supplier, customer and product or service, after company identities are reconciled | 2026-09-11 | the graph scope | Counted when the snapshot was created, after duplicate relationships were removed. |
| Export-control designations | 3,435 | recorded export-control designations, with Federal Register sources | 2026-09-11 | the substrate snapshot | Current count agrees (3,435) |
| Designation-to-company matches | 176 | active matches between designations and company records; a designation can have several matches | 2026-09-11 | the substrate snapshot | Match records have restricted access. This total is counted when the snapshot is created. |
| Sovereign programs | 36 | national or regional government compute programs in the curated list | 2026-09-11 | the programme list | Counted from the current curated program list. |
| Public query tools | 55 | tools available to the public Query Builder; excludes administrative tools | current tool list | the tool list | Counted from the Query Builder’s public tool list. |
Limitations
Coverage depends on disclosure. Public-company filings can identify subsidiaries and ownership relationships that remain unknown for private or state-owned operators; missing reports also limit facility-capacity estimates.
Power and capital cost support capacity estimates only through assumptions about the hardware they serve. Hardware inventories reduce that uncertainty, but utilization and training duration still affect estimates of how much compute a facility can supply.
A reported supply relationship rarely establishes how much production would be lost if it failed. The cascade model uses default supply shares where values are missing and analyst-selected propagation rates; estimated severity depends on those assumptions.
Built by
David Gringras, physician (Edinburgh) and law graduate (University of Law), Frank Knox Fellow at Harvard (MPH Health Policy, 2026; cross-registered at MIT, Harvard Law School, and the Kennedy School). Expert panelist and co-author on the Evals-Consensus.AI Delphi study of AI evaluation practice; supported by a BlueDot Impact career transition grant. Previously Evaluations and Collaborations Lead on the FATF-to-AI governance translation project at Arcadia Impact / The Future Society, and project supervisor at Orion AI Governance on evaluation-independent governance mechanisms.
Send corrections or methodological criticism to the address above. Published corrections are listed at Corrections.