OpenAI Puts Stargate UK Data Center on Ice, Citing Energy Costs and Regulatory Red Tape
Available in: 中文
OpenAI has paused plans for a major data center in the UK as part of the Stargate Project, citing high energy costs and regulatory complexity as primary obstacles. The Register reports that the dec...
OpenAI Puts Stargate UK Data Center on Ice Over Energy Costs and Regulatory Hurdles
OpenAI has paused plans for a major data center in the UK as part of the Stargate Project, citing high energy costs and regulatory complexity as primary obstacles. The Register reports that the decision reflects the growing difficulty of building AI infrastructure in Europe.
The Stargate Project
The Stargate Project is a billion joint venture between OpenAI, SoftBank, and Oracle to build massive AI infrastructure globally. UK was considered a key European location due to:
- English-speaking workforce: Skilled AI and engineering talent
- Financial sector: Proximity to London financial services industry
- Regulatory alignment: UK GDPR regime post-Brexit
- Political support: UK government expressed strong interest in hosting AI infrastructure
Why the Pause
Key factors behind the decision:
- Energy costs: UK electricity prices are among the highest in Europe, making data center operations expensive
- Grid connection delays: Securing grid capacity for large data centers can take years in the UK
- Regulatory complexity: Planning permissions, environmental assessments, and energy regulation create lengthy approval processes
- US alternatives: More favorable conditions in the US where energy is cheaper and regulations are more streamlined
What This Means for the UK
The UK AI strategy takes a significant hit:
- Lost investment: Billions in potential infrastructure investment redirected elsewhere
- Talent risk: AI companies may locate infrastructure where talent follows
- Competitive disadvantage: Other European countries may also lose out to US and Middle East locations
- Political embarrassment: UK government has heavily promoted itself as an AI leader
The Energy-Data Center Paradox
This case illustrates the fundamental challenge:
- AI needs massive power: Training frontier models requires gigawatts of electricity
- Europe expensive energy: Energy costs are 2-3x higher than in the US or Middle East
- Climate commitments: European nations have stricter emissions targets that complicate data center siting
- Grid capacity: European grids were not designed for the industrial-scale power demands of AI
Global Data Center Race
Other locations are winning the AI infrastructure competition:
- US: Abundant energy, favorable regulation, tech ecosystem
- Middle East (UAE, Saudi Arabia): Cheap energy, government backing, new nuclear deals
- Southeast Asia: Growing capacity, competitive costs, geopolitical diversification
- Japan: Technology base, nuclear restarts providing clean power
Source: The Register — April 9, 2026
← Previous: Maine Becomes First US State to Ban Major New Data Centers: Climate and Grid Concerns Drive Landmark LegislationNext: Research-Driven Agents: What Happens When AI Agents Read Before They Code →
0