---
id: "claim-energy-dictates-generative-ai"
type: "claim"
source_timestamps: ["§ Thinking About AI Capability on a National Scale", "§ How to Develop a Country-Level AI Strategy"]
tags: ["infrastructure", "energy", "generative-ai"]
related: ["framework-national-ai-capability", "action-scout-locations-by-need", "entity-canada", "prereq-generative-vs-applied-ai"]
confidence: "high"
testable: true
speakers: ["Yasuhiro Yamakawa", "Thomas H. Davenport"]
sources: ["futures"]
sourceVaultSlug: "hbr-seg-futures"
originDay: 2
articleStem: "hbr-cl-94-ai-strategy-beyond-us-china"
sourceUrl: "https://hbr.org/2025/12/your-ai-strategy-needs-to-expand-beyond-the-u-s-and-china"
sourceTitle: "Your AI Strategy Needs to Expand Beyond the U.S. and China"
---
# Energy availability dictates the location of generative AI model training

**Claim:** Energy availability dictates the location of generative AI model training.

**Confidence: high · Testable: yes**

Cutting-edge development of large generative AI models requires massive amounts of energy to power data centers. Countries with surplus or easily expandable energy therefore offer a distinct competitive advantage for this specific type of AI development. The authors explicitly cite **France** (nuclear output), [[entity-canada]] (hydroelectric), **Sweden**, and **Norway** (hydroelectric) as prime locations for companies needing large amounts of electricity for AI-focused data centers.

This is why understanding [[prereq-generative-vs-applied-ai]] matters — energy is a *training* constraint, not an application constraint — and it is one factor in the [[framework-national-ai-capability]]. Operationally it drives [[action-scout-locations-by-need]].

**Enrichment assessment:** Directionally supported, but "dictates" is too strong — energy is a *major but not sole* determinant. Training frontier models needs hundreds of MWh to GWh; hyperscalers place compute-intensive data centers in regions with abundant low-cost, low-carbon power (France's nuclear base; Canada/Sweden/Norway's hydro, cool climates, political stability). IEA and cloud-provider analyses flag electricity availability and grid decarbonization as key constraints. **But** network latency to markets, talent, tax/subsidy regimes, political risk, and data-sovereignty laws also drive siting; firms run a *portfolio*. Verdict: **Mostly supported** if rephrased as "strongly shapes" rather than "dictates."


## Related across articles
- [[claim-ai-bottleneck-electricity]]
- [[concept-new-ai-triad]]
- [[action-secure-energy]]