AI BriefingBuilding AI in America
Key point
Anthropic emphasized that the United States needs large-scale investment in computing and power infrastructure to maintain AI leadership.
Details
For the United States to lead the world in AI, massive investment in computing power and electricity is essential. Anthropic has released policy recommendations to help train world-class AI models within the United States and deploy them across all industries.
Training frontier AI models requires a stable and continuous power supply. Anthropic expects the need for 2GW of data centers by 2027 and 5GW by 2028, and projects that total U.S. demand for frontier AI training will reach 20-25GW by 2028.
For the U.S. AI sector to maintain global leadership, at least 50GW of power capacity must be secured by 2028. China is currently building out energy infrastructure much faster than the United States, so the U.S. needs to take active measures such as accelerating permitting processes and supporting grid expansion.
Key policy recommendations are as follows:
- Accelerate NEPA (National Environmental Policy Act) reviews and permitting processes
- Utilize federal land as sites for AI infrastructure construction
- Support grid interconnection and transmission projects
- Support the use of energy technologies such as next-generation geothermal and advanced nuclear technology
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
NVIDIA Expands AI Infrastructure Financing Guarantees... Strengthening Role as 'Central Bank of AI'