
Goldman Sachs discusses "Trump 2.0" energy policy: AI consumes electricity, more nuclear power, fewer electric vehicles

Last week, the executive order issued by Trump included a reduced emphasis on electric vehicles and offshore wind energy, as well as support for nuclear energy. Analysts believe that with the growing demand for electricity from artificial intelligence and data centers, large technology companies are willing to pay a Green Reliability Premium to support nuclear energy or other clean and reliable power sources
In the "Trump 2.0" era, insufficient infrastructure investment and rising electricity demand will further promote the development of nuclear power.
Recently, Goldman Sachs analysts Brian Singer, Brendan Corbett, and others discussed the "energy emergency" executive order issued by Trump, as well as the private sector's announcement of the "Stargate" artificial intelligence infrastructure project in their report.
Goldman Sachs stated that although Trump's executive order froze all funding allocations from the Inflation Reduction Act (IRA), Goldman Sachs remains bullish on several sustainable development themes, which include 1. Reliability of energy, electricity, and water supply; 2. Innovations in energy, land, and resource utilization efficiency; 3. Increased demand from sustainable investors for AI and automation.
Goldman Sachs explained that with the growth of artificial intelligence and electricity demand from data centers, large technology companies are willing to pay a Green Reliability Premium to support nuclear energy or other clean and reliable power sources.
In addition, sustainable investors increasingly recognize that AI and automation are essential for addressing the labor challenges posed by the aging population in developed economies, and they are becoming increasingly important in promoting retraining and education.
Goldman Sachs emphasized that the executive orders related to energy and artificial intelligence issued by Trump last week align with Goldman Sachs' previous expectations, including:
- Prioritizing the acceleration of permitting processes for energy, electricity, and mineral development in the U.S.;
- Reducing emphasis on electric vehicles and offshore wind energy;
- Suspending further IRA loans and grants;
- Supporting nuclear energy;
- Supporting infrastructure, artificial intelligence, and data centers.
Surge in AI and Data Center Electricity Demand Accelerates Nuclear Energy Promotion
Goldman Sachs expects that by 2030, global electricity demand from data centers will increase by 160%-165% compared to 2023, with data center electricity demand rising from 1%-2% of total global electricity demand to 3%-4%.
However, this growth may face five potential constraints:
- AI server shipments may be limited by data center capacity;
- Data center capacity may be limited by electricity infrastructure;
- Electricity infrastructure may be limited by low-carbon options/costs;
- Next-generation AI chips may lead to either a decrease or increase in overall electricity demand;
- AI server demand may be constrained by AI outcomes/innovations.
In response, Goldman Sachs analysts believe that rising electricity demand, historically insufficient infrastructure investment, and increasing temperatures/extreme weather events will continue to drive the market's demand for reliability investments, as both infrastructure renewal and reinforcement require reliability investments, and adaptability will become an increasingly important theme regardless of climate.
Therefore, Goldman Sachs stated that a more significant shift is occurring in the "Trump 2.0" era:
"Accelerating nuclear power expansion and reducing U.S. electric vehicle incentives may not undermine overall green capital expenditure, while potentially leading to a long-term net reduction in carbon emissions."