Critical resources & minerals
Measures control over the materials that make AI chips, batteries and infrastructure possible.
Publication · 2026 Edition
AI Capital Index
Artificial intelligence is not immaterial: the index ranks nations by their control of the physical, financial, human and geopolitical foundations on which it depends.
Key figures to cite
- ~74 %
- of global cobalt (DRC)
- ~60 %
- of global coltan (DRC)
- 2nd
- largest copper producer (DRC)
How to cite this study: Lumen Global Inc., AI Capital Index 2026, https://lumenglobalinc.com/en/insights/ai-capital-index
The thesis
Power in AI begins long before the model.
Artificial intelligence rests on a concentrated physical and institutional architecture: critical minerals, energy, compute, talent, financial capital and governance. These assets are scarce, geographically concentrated and increasingly contested.
The AI Capital Index measures who controls these foundations, link by link, from mine to model. It distinguishes a country’s initial endowment from its ability to convert that position into industry, influence and enduring value.
The index consolidates two earlier Lumen research programmes — the Global AI Sovereignty Index and the AI Value Chain Index — into a common architecture designed to compare nations and inform public policy and capital allocation.
- ~74 %
- of global cobalt
- ~60 %
- of global coltan
- 2nd
- largest copper producer
Index architecture
What the index measures
Energy
Assesses the availability, reliability and cost of the electricity required for intensive computing.
Compute & digital infrastructure
Maps data centres, semiconductors, networks and sovereign computing capacity.
Human capital
Examines education, research, attraction and retention of critical expertise.
Innovation & capital
Measures the depth of finance, applied research and the entrepreneurial ecosystem.
Industrial transformation
Assesses the ability to convert upstream assets into production, adoption and domestic value added.
Governance
Examines institutions, rules, data control and the state’s capacity to shape the value chain.
Detailed weights and indicators are published in the report.
- 01Extraction
Cobalt and copper (DRC) · Bauxite (Guinea) · Phosphate (Morocco)
- 02Refining and processing
From ore to industrial material
- 03Components
Chips, batteries, printed circuits
- 04Infrastructure and energy
Data centres, networks, electricity
- 05Models and uses
Where conventional rankings stop
Ranks and scores by link: reserved for the full edition
Country case study
Democratic Republic of the Congo
The DRC holds an exceptional upstream position in the mineral chain behind the AI stack. It accounts for about 73% of world cobalt mine production, ranks as the world's second-largest copper producer behind Chile, supplies roughly half of the world's tantalum, the metal extracted from coltan, and holds, notably at Manono, lithium resources estimated at 3 million tons.
Yet most refining, processing, finance, intellectual property and industrial value are concentrated elsewhere. Resource endowment does not automatically translate into economic or strategic power.
The case study traces the gap between upstream control and downstream value capture. It identifies where the Congolese position is strong, where it dissipates and which policy choices could alter that trajectory.
Sources: U.S. Geological Survey, Mineral Commodity Summaries 2026 (cobalt, copper, tantalum, lithium), 2025 estimates; DRC Ministry of Mines, CTCPM, mining statistics.
Methodology
A comparable, revisable and transparent architecture.
Sources
Identified public indicators, institutional datasets and country-level documentary checks.
Weighting
Seven weighted pillars aggregated into a comparative Top 25 index.
Limitations
Data gaps, time lags and limits to comparability are disclosed alongside the results.
Review
Indicators, weights and country cases are reviewed for each annual edition.
Publication · 2026 Edition
Download the report
Receive the reference edition on control and value capture across the global artificial-intelligence value chain.